Product Liability and Microbial Foodborne Illness

Home , Adulterated food

Product Liability and Microbial Foodborne Illness. By Jean C. Buzby, Paul D. Frenzen, and Barbara Rasco. Food and Rural Economics Division, Economic Research Service, U.S. Department of Agriculture. Agricultural Economic Report No. 799.

Abstract This report examines how product liability law treats personal injuries attributed to microbially contaminated foods. The risk of lawsuits stemming from microbial foodborne illness and the resulting court-awarded compensation may create economic incentives for firms to produce safer food. It is not known how many consumers seek compensation for damages from contaminated foods because information about complaints and legal claims involving foodborne illness is not readily accessible, especially for cases that are settled out of court. Reviewing the outcomes of 175 jury trials involving foodborne pathogens, the analysis identifies several factors that influence trial outcomes, while noting that the awards won by plaintiffs tend to be modest.

Keywords: foodborne illness, food safety, jury verdict, personal injury, product liability

Acknowledgments We thank David Isenburgh, Peter Machare, and Marilynn Graham for their assistance with Lexis-Nexis and Westlaw databases and to Jane Allshouse for her help with our database. We also thank Bruce T. Clark, Katherine Ralston, Carolyn Rogers, Donna Rosenbaum, and Vern Walker for their helpful review comments. Tom McDonald edited the manuscript. Susan DeGeorge designed the cover art.

1800 M Street Washington, DC 20036-5831 April 2001 Contents Page Summary ...... iv

Chapter 1: Introduction ...... 1

Chapter 2: Characteristics of Microbial Foodborne Illness Relevant to Litigation ...... 3 The Pathology of Foodborne Illness ...... 3 Causes of Foodborne Illness ...... 3 Food Handling Errors by Firms ...... 5 Food Handling Errors by Consumers ...... 5 Epidemiological and Legal Perspectives on the Causes of Foodborne Illness ...... 6 Characteristics of Foodborne Illness ...... 7 Economic Costs of Foodborne Illness ...... 7

Chapter 3: The Intersection between Liability Law and Economics and Its Relevance to Foodborne Illness Litigation ...... 8 Overview of the Incentive System To Provide Food Safety ...... 8 Intersection Between Law and Economics ...... 8 Insurance and Foodborne Illness ...... 9 Transaction and Information Costs ...... 11 Transaction Costs ...... 11 Information Costs ...... 11

Chapter 4: Analysis of Jury Verdict Data for Foodborne Illness ...... 13 U.S. Foodborne Illness Court Data ...... 13 Findings ...... 14 Time to Legal Resolution ...... 14 Frequency and Size of Awards ...... 15 Court Awards by Severity Category ...... 15 Court Awards by Implicated Pathogen ...... 16 Court Awards by Implicated Food ...... 16 Court Awards by Type of Defendant ...... 17 Other Information Provided by Court Data ...... 18 Key Findings ...... 18 Multivariate Analyses of the Court Data ...... 19 Lawsuit Characteristics ...... 19 Plaintiff Characteristics ...... 19 Defendant Characteristics ...... 20 Design of Analysis ...... 20 Independent Variables and Their Hypothesized Effects ...... 20 Regression Results and Interpretation ...... 22 Discussion ...... 23

Chapter 5: Summary ...... 24

References ...... 28

ii Appendix: Product Liability Law As It Applies to Foodborne Illness ...... 34 The Legal Process ...... 34 U.S. Product Liability Law Is State Law ...... 35 Causes of Action ...... 35 Strict Product Liability ...... 36 Negligence ...... 36 Breach of Express or Implied Warranty ...... 38 Common Defenses in a Product Liability Lawsuit ...... 39 Legal Compensation for Foodborne Illness ...... 40 Defendant Costs ...... 40 Discussion ...... 40 Class Action Suits ...... 40

Tables Page 1. Estimated annual foodborne illnesses, hospitalizations, and deaths, due to selected pathogens, United States, 1999 ...... 5 2. Compensation for consumer plaintiffs in foodborne illness lawsuits decided by jury verdicts, 1988-97 ...... 15 3. Compensation in foodborne illness court cases by severity category, 1988-97 ...... 16 4. Foodborne pathogens, toxins, or illnesses involved in foodborne illness lawsuits decided by jury verdicts, 1988-97 ...... 16 5. Compensation in foodborne illness court cases by pathogen category, 1988-97 ...... 17 6. Food items involved in foodborne illness lawsuits decided by jury verdicts, 1988-97 ...... 17 7. Compensation in foodborne illness court cases by food category, 1988-97 . . . . . 18 8. Defendants in foodborne illness court cases by firm type, 1988-97 ...... 18 9. Definitions and mean values of independent variables ...... 21 10. Multivariate analyses of foodborne illness jury verdicts ...... 22

Appendix table 1. Three causes of action for product liability cases ...... 37

iii Summary

Food firms, such as manufacturers, retailers, and restaurants, have economic incentives to produce safer food in order to avoid foodborne illness lawsuits and the potential compensation that they may have to pay to ill people and their families. Lawsuits would seem to provide important feedback to these firms about how much they should invest in food safety. However, high transaction and information costs, combined with the structure of the legal system, limit the effectiveness of litigation for compensating ill consumers and providing firms with signals to produce safer food. This report reviews earlier work on the economic and legal aspects of foodborne illness and examines how product liability law treats personal injuries attributed to contaminated food products. Data on jury trials involving personal injuries allegedly due to foodborne pathogens during 1988-97 were analyzed using multivariate methods to identify the factors related to trial outcomes and the size of damage awards. Among other findings: Plaintiffs are unlikely to receive awards in foodborne illness jury trials. Relatively few foodborne illnesses are compensated either through jury awards or out-of-court settlements. Of our sample of 175 foodborne illness lawsuits resolved in court during 1988- 97, 31.4 percent resulted in some compensation paid by firms. Plaintiffs were more likely to win jury trials if they could link their illness to a specific pathogen, and more severe illnesses tended to result in higher awards. Multivariate analyses highlight the importance of plaintiffs being able to link their illness to a specific foodborne pathogen in order to prevail in court. Expected monetary compensation from a foodborne illness lawsuit provides a limited incentive to pursue litigation. The median award by juries for injuries due to pathogen- contaminated food products was $25,560 (1998 dollars). Plaintiffs seldom receive all of an award because part of the award (typically one-third or more) pays legal fees and court costs. Foodborne illness costs are shared by many sectors of the economy, which limits incentives to firms to produce safer food. Much of the costs of illness borne by people who become ill (and/or their families) are not reimbursed by food firms responsible for an illness. Rather, an ill consumer, his or her relatives, other parties (such as employers, private health insurers, and taxpayers), or some combination of these bear the costs. Legal incentives probably work better in outbreaks and less well for sporadic cases. Mass outbreaks have greater potential to damage firms, both in terms of financial damages and of damaging a firms or a products reputation. Whether foodborne illness litigation will become more common in the future is unclear. However, class action lawsuits may become more common for outbreaks, which result in many mild and similar illnesses, particularly as identification and documentation of outbreaks improve, as legal expertise in this area grows, and as media coverage of successful class action suits involving consumer products accumulates. In short, the legal system provides limited incentives for firms to produce safer food. Firms responsible for microbial contamination compensate relatively few people who experience foodborne illnesses. The product liability system provides firms with incentives to control hazards in food primarily when the hazards are easily identifiable, a foodborne illness can be traced to firms, and ill people or their families are compensated by the firms responsible for the contamination. iv Product Liability and Microbial Foodborne Illness

Jean C. Buzby, Paul D. Frenzen, and Barbara Rasco

Chapter 1

Introduction

Food products may cause human illness if they contain Financial compensation for injuries due to contaminat- microbial pathogens such as bacteria, parasites, fungi, ed food products is important because foodborne ill- or viruses. Foodborne illness is relatively common in ness imposes substantial economic costs on society. the United States despite intensive efforts by Govern- The annual medical costs, productivity losses, and ment agencies and private firms to ensure that food costs of premature deaths due to five major foodborne products are safe. The U.S. Centers for Disease pathogens are estimated to be $6.9 billion (Crutchfield Control and Prevention (CDC) has estimated that 76 and Roberts, 2000). This estimate represents only a million foodborne illnesses occur each year in the fraction of the total costs due to foodborne illness, United States, resulting in 325,000 hospitalizations and which include some costs, such as pain and suffering, 5,000 deaths (Mead et al., 1999). Pathogen-contami- that are difficult to quantify and other costs, such as nated foods consequently represent an important cause public health expenditures on foodborne disease, that of unintentional injury and death. In fact, contaminat- are often overlooked. ed food products caused more deaths each year than Economic theory suggests that firms that make or dis- the combined totals of all 15,000 products regulated by tribute food products will invest fewer resources in the U.S. Consumer Product Safety Commission; these reducing disease-causing contamination if they expect products caused 3,700 accidental deaths in 1996 (U.S. not to pay for injuries due to contaminated products. Consumer Product Safety Commission, 1998). When firms escape paying compensation, the costs of This report focuses on foodborne illnesses caused by injuries are borne instead by the consumers who food-handling errors by firms and the subsequent law- become ill or are shifted to other parties, such as health suits triggered by those errors. Under U.S. product lia- insurers and employers that provide sick leave bene- bility law, people harmed by unsafe products (includ- fits. In contrast, if firms expect to bear the costs of ing foods contaminated by microbial pathogens) can injuries due to contaminated products, they will likely take legal action to claim money damages for their invest more resources in reducing contamination. injuries. Product liability law specifies when firms are Lawsuits by consumers injured by foodborne liable for injuries due to their products and are required pathogens are one signal for firms to spend more for to pay compensation to injured persons or their sur- food safety in order to reduce compensation costs. If vivors. In the case of contaminated foods, product lia- most firms currently escape paying compensation, bility is a seemingly powerful mechanism to compen- increased corporate investments in food safety should sate consumers for economic losses due to foodborne lower the overall incidence of foodborne illness and illness, while simultaneously encouraging firms to pro- liability costs and result in a more optimal sharing of vide safer food products. food safety costs between firms and consumers.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 1 Lawsuits by consumers to recover damages due to This report examines how foodborne illness is handled foodborne illness can affect the behavior of firms that under U.S. product liability law. We review liability make or distribute food products. The magnitude of law for injuries due to food products contaminated by this effect is unknown, however, because information microbial pathogens, and we investigate the incidence about litigation involving injuries due to food products of litigation and the characteristics of cases tried in contaminated by microbial pathogens is scarce. Firms State and Federal courts. Most foodborne illnesses do (or their product liability insurers) generally prefer to not result in litigation. Furthermore, only a third of the resolve consumer complaints about foodborne illness consumers who go to court receive financial compen- outside the courtroom, where they can keep compensa- sation for their injuries. These findings suggest that tion payments confidential, and avoid or reduce the direct impact of court decisions on firms is small, adverse publicity about their products. Some lawsuits although few if any firms are likely to ignore the result in trials, but court statistics do not distinguish potential legal consequences of selling contaminated cases involving contaminated foods from other product food products that could cause illness or death. liability cases. Court decisions about liability for food- Chapter 2 discusses the characteristics of foodborne borne illness may also appear inconsistent because illness that may influence litigation involving injuries product liability law is complex and the disposition of due to foodborne pathogens. Chapter 3 provides an court cases for contaminated food products is evolving. overview of the general concept of product liability For example, liability currently varies according to the and its economic impacts. For readers interested in type of food and pathogen involved in causing illness, Federal and State law regarding liability for foodborne reflecting previous court decisions and judges and illness, the appendix provides a comprehensive back- juries assumptions about consumer awareness of the ground. Chapter 4 analyzes U.S. jury verdict data on safety of particular foods, as well as existing laws and foodborne illness lawsuits for 1988-97. regulations.

2 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA Chapter 2

Characteristics of Microbial Foodborne Illness Relevant to Litigation

The Pathology of Foodborne Illness nesses result in secondary complications that may become chronic health problems (Lindsay, 1997). The More than 40 different foodborne microbial pathogens best-known complications associated with foodborne are known to cause human illness, including bacteria, pathogens include reactive arthritis, hemolytic uremic parasites, viruses, fungi, and their toxins (CAST, 1994, syndrome (characterized by kidney failure), and pp. 11-15). The ecology of pathogens varies. Some Guillain-Barré syndrome (characterized by neuromus- pathogens, such as Listeria monocytogenes, are perva- cular paralysis). Table 1 provides estimates of the sive in the natural environment and may contaminate annual foodborne illnesses, hospitalizations, and deaths food during production or distribution. Others have for some of the most common or deadly foodborne found new ecological niches, such as Salmonella pathogens. serotype Enteritidis in eggs. Although most of the estimated 76 million annual Several pathogens were recognized only recently as a foodborne illnesses in the United States are relatively cause of foodborne illness (Tauxe, 1997). Some food- mild and self-limiting, at least 325,000 of these illness- borne pathogens have not yet been scientifically identi- es are serious enough to result in hospitalization or fied. The CDC has estimated that these elusive, death (Mead et al., 1999). Foodborne illnesses account unknown pathogens account for 81 percent of the for about 1 of every 100 U.S. hospitalizations and 1 of foodborne illnesses in the United States (Mead et al., every 500 U.S. deaths.2 The annual number of deaths 1999). These unknown pathogens probably account to due to food products contaminated by microbial some extent for epidemiologists inability to identify pathogens is much smaller than the number of deaths the pathogens that caused over two-thirds of the 2,800 associated with certain other products, notably tobac- mass outbreaks of foodborne illness reported to the co, medical drugs, and alcoholic beverages. However, CDC during 1993-97 (Olsen et al., 2000).1 contaminated foods are responsible for many more accidental fatalities than some products commonly per- The illnesses caused by foodborne pathogens vary ceived as dangerous, including firearms, industrial greatly in severity, duration, and clinical manifesta- machinery, and explosives. tions. Most foodborne illnesses are not severe or pro- longed and are limited to brief episodes of diarrhea, nausea, or other acute gastrointestinal symptoms. A Causes of Foodborne Illness small proportion of foodborne illnesses are severe or fatal, however. The CDC has estimated that 0.43 per- Epidemiological investigations of foodborne illnesses cent of U.S. foodborne illnesses require hospitaliza- provide some information about the specific pathogens tion, while 0.01 percent result in death (Mead et al., and foods that caused illness. Animal products such as 1999). The most severe acute illnesses associated with meat, poultry, seafood, dairy products, and eggs are the foodborne pathogens include complications such as foods most likely to cause outbreaks of human illness septicemia (infection of the bloodstream), localized in the United States (CAST, 1994, p. 32). In recent infections of other organs, and spontaneous abortion in years, the variety of foods associated with foodborne pregnant women. About 2-3 percent of foodborne ill- illness has increased (Tauxe, 1997). Some examples include salami, lettuce, bean sprouts, and raspberries. Epidemiological investigations of foodborne illnesses also identify the kinds of errors in food production, 1 CDC data on foodborne disease outbreaks define an outbreak as an incident in which two or more persons experi- enced a similar illness after ingestion of a common food, and epidemiologic analysis implicated a food as the source 2 The United States had 31.1 million community hospital of the illness. There are two exceptions, botulism and chem- admissions in 1996 (U.S. Census Bureau, 1999) and 2.3 ical poisoning, for which one case constitutes an outbreak. million deaths in 1998 (Murphy, 2000).

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 3 distribution, or preparation that allow microbial borne illness (Powell, 1999). Most information about pathogens to contaminate food.3 Some food-handling the errors that caused illness is derived from epidemio- errors introduce pathogens into uncontaminated food. logical investigations of foodborne illness outbreaks by Other errors permit the pathogenic organisms already State and local public health agencies. However, present in raw food to survive or multiply to dangerous investigated outbreaks account for only a small and levels in prepared food. Potential errors include: nonrepresentative share of all foodborne illnesses for several reasons. Public health agencies are more likely the use of contaminated raw food, l to learn about outbreaks affecting many people than l cross-contamination of prepared food by contaminat- about sporadic cases of illness affecting only one per- ed raw food, son, although sporadic cases are much more frequent than outbreak cases (Bean et al., 1990). Public health l poor personal hygiene by infected food handlers, agencies are also more likely to learn about certain l inadequate cleaning of equipment, kinds of outbreaks than others, notably large outbreaks and outbreaks involving restaurants or severe illness l inadequate cooking or reheating, (Bean et al., 1996). Finally, public health agencies l improper holding temperatures, have limited resources and do not thoroughly investigate or report every known outbreak to the CDC l cooling food too slowly after heating, (Berkelman et al., 1994; Bean et al., 1996). l eating food too long after preparation, The most recent CDC summary of foodborne illness outbreaks indicates that public health agencies reported l insufficient fermentation, acidification, salting, or an annual average of 550 outbreaks resulting in 17,200 sweetening during processing (Bryan et al., 1997). foodborne illnesses during 1993-97 (Olsen et al., The most common cause of recent mass outbreaks of 2000). Reported outbreaks consequently accounted for foodborne illness reported to the CDC was improper only about one of every 4,000 foodborne illnesses in holding temperatures, but many outbreaks involved the United States. About two-fifths of the outbreak more than one error (Olsen et al., 2000). reports sent to CDC did not identify the food-handling errors that caused illness (Olsen et al., 2000). Both food firms and consumers make food-handling errors that result in foodborne illness.4 Many illnesses Epidemiological case-control studies of sporadic cases are attributable to sequential errors by firms and con- of foodborne illness also provide information about sumers. Sequential errors occur when consumers food-handling errors that cause illness. For example, a improperly handle foods that were initially contaminat- case-control study revealed that sporadic E. coli ed by microbial pathogens during commercial produc- O157:H7 infections are associated with eating under- tion or distribution. For example, a meatpacking plant cooked hamburgers (Slutsker et al., 1998). However, may fail to prevent ground beef from being contami- case-control studies have important limitations, notably nated by Salmonella bacteria, and consumers may sub- their reliance on consumers self-reports about food sequently undercook hamburgers made from the handling. ground beef, causing those who eat the hamburgers to Despite the lack of information about the specific become sick. errors that caused most foodborne illnesses, some The proportion of foodborne illnesses due to separate experts have concluded that most illnesses are attribut- food-handling errors by firms and consumers is able to food-handling errors by consumers (Scott and unknown due to the limitations of the data on food- Sockett, 1998). Many consumers engage in unsafe food-handling practices, and some consumers prefer to eat risky foods. Examples include placing cooked hamburgers back on a plate that contains raw meat 3 For simplicity, we use the term food-handling errors to juices, and eating Caesar salad made with raw eggs. include errors in food production, distribution, or prepara- To conclude that most foodborne illnesses are due to tion. consumers alone, however, ignores ample evidence 4 The discussion of firms in this report covers noncommer- that firms also make food-handling errors resulting in cial organizations such as schools and churches as well as foodborne illness. commercial firms, because noncommercial organizations may also handle food and be held liable for injuries due to foodborne pathogens.

4 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA Food-Handling Errors by Firms breaks attributed to food prepared by firms has increased over time, rising from 63 percent in 1973-75 The most recent CDC summary of reported foodborne to 78 percent in 1993-97, although the reasons for the disease outbreaks covering the 1993-97 period (Olsen increase are unknown (Bean and Griffin, 1990; Olsen et al., 2000) provides a sample of food-handling errors, et al., 2000). Although food-handling errors by firms most of which were never reported in the mass media. were involved in most reported outbreaks, little infor- The sample includes information about the food prod- mation is available about the role of firms in causing uct, the place where the food was eaten, and the food- either unreported outbreaks or sporadic cases of food- handling errors responsible for causing illness. The borne illness. 2,751 outbreaks included in the sample involved a wide variety of places and food products, indicating that food-handling errors were not restricted to a few Food-Handling Errors by Consumers error-prone firms or risky foods. The high frequency of risky food-handling practices The CDC summaries of reported foodborne disease and food preferences among consumers suggests that outbreaks reveal that most reported outbreaks were many illnesses are due at least in part to consumer caused by food-handling errors by firms. During behavior. In addition to the risky food-handling prac- 1993-97, nearly 78 percent of the outbreaks with infor- tices already mentioned, other common risky practices mation about the place where contaminated food was include undercooking raw meat and poultry, and eaten occurred in a commercial or institutional estab- preparing salad and raw meats with the same utensils lishment, while only 22 percent occurred in a private and cutting boards. The 1998 FDA Food Safety residence (Olsen et al., 2000). The proportion of out- Survey of U.S. adults found that 4 percent ate raw

Table 1—Estimated annual foodborne illnesses, hospitalizations, and deaths due to selected pathogens, United States, 1999

Disease or agent Illnesses Hospitalizations Deaths Comment

------Number ------Bacterial Campylobacter spp. 1,963,141 10,539 99 A small percentage of people develop Guillain-Barré Syndrome. Clostridium perfringens 248,520 41 7 Usually causes mild gastrointestinal distress lasting only a day. Escherichia coli O157:H7 62,458 1,843 52 Usually a mild gastrointestinal illness, but severe complications such as bloody diarrhea and kidney failure may develop (e.g., hemorrhagic colitis and hemolytic uremic syndrome (HUS)). Listeria monocytogenes 2,493 2,298 499 Women infected with Listeria during pregnancy may transmit the infection to the fetus, possibly leading to stillbirths or babies born with mental retardation. Salmonella, nontyphoidal 1,341,873 15,608 553 Relatively mild and common. Staphylococcus foodborne illness 185,060 1,753 2 Characterized by severe nausea, vomiting, cramps, and diarrhea. Vibrio cholerae, toxigenic 49 17 0 Causes epidemic cholera. Vibrio vulnificus 47 43 18 Fatality rate is about 50 percent in people with chronic liver disease. Parasitic Toxoplasma gondii 112,500 2,500 375 Infection may be transmitted to fetuses, possibly leading to stillbirths or babies born with birth defects ranging from hearing or visual impairments to mental retardation. Source: Data from Mead et al., 1999.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 5 steak tartare, 37 percent ate raw eggs or products with due to illness, which depends on other factors in addi- raw eggs, and 12 percent ate raw oysters (Fein and tion to responsibility for the food-handling errors that Riggins, 1998). caused illness. For example, although consumers and food firms may share responsibility for causing food- Consumers should follow proper food safety practices borne illness under the law, lawyers for food firms because raw foods might be contaminated by microbial generally avoid blaming consumers for making food- pathogens. Proper handling, cooking, and storage pro- handling errors because jurors also make the same cedures will eliminate most pathogens. errors at home and would likely be sympathetic to consumers accused of such errors (Clark, 2000).5 The law Epidemiological and Legal Perspectives on also imposes different standards of conduct on food the Causes of Foodborne Illness firms than on consumers because food firms are expected to know about the risks of foodborne The epidemiological perspective on foodborne illness pathogens and to take effective measures to prevent is separate from the legal perspective discussed more pathogen contamination (Clark, 2000). fully in the next chapter. From an epidemiological perspective, the example of Salmonella-contaminated ground beef mentioned earlier was due to errors by both the meatpacking plant and consumers who prepared hamburgers. In essence, the epidemiological perspective focuses on exactly how a food product 5 Any firm that raises the issue of consumer handling errors became contaminated by a pathogen and caused must also deal with the issue of why the firm did not warn human illness. In contrast, the legal perspective on the consumer about potential risks due to pathogen contami- foodborne illness focuses on liability for the damages nation of its food products. The legal concept of failure to warn is discussed further in the appendix.

Figure 1 Usual incubation period ranges for select foodborne diseases

Hepatitis A Listeria monocytogenes1

Yersinia enterocolitica

E. coli O157:H7 Campylobacter Shigella Cyclospora cayetanensus Clostridium botulinum Salmonella, nontyphoidal Clostridium perfringens

Vibrio parahaemolyticus Staphylococcus aureus

22881420 26 5 11 14 17 20 23 26 29 32 35 38 41 44 47 50

Hours Days

1Invasive form, incubation period for diarrheal disease unknown. Source: Data on the "usual" incubation period obtained from the Centers for Disease Control and Prevention, "Surveillance for Foodborne- Disease Outbreaks--United States, 1988-1992." MMWR 45,SS-5 (Oct. 25, 1996): 58-66.

6 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA Characteristics of Foodborne Illnesses diet-related factors, such as consumption of antacids and nutritional deficiencies (CAST, 1994, p. 27). Several characteristics of foodborne illnesses may Foodborne illnesses tend to have the most severe con- affect efforts to determine liability for the injuries due sequences in children, the elderly, and the immuno- to these illnesses. These characteristics include the compromised (i.e., AIDS and cancer patients). incubation period, mass outbreaks of illness, a persons susceptibility to foodborne infections, and medical Medical Practice Patterns. Most foodborne illness practices. patients never receive a definitive medical diagnosis, either because they did not seek medical care or Incubation Period. Many foodborne pathogens do not because their physician did not test for the pathogen cause symptoms of acute illness until several days that caused the illness. For example, the Foodborne after the contaminated food is eaten, although the aver- Diseases Active Surveillance Network (FoodNet) age incubation period before the onset of acute symp- found that only 12-15 percent of people who experi- toms varies considerably by pathogen (fig. 1). enced acute diarrhea (the most common symptom of Symptoms due to the toxins produced by foodborne illness) saw a doctor about their illness, and Staphylococcus aureus appear within 2-4 hours, while only about a fifth of diarrhea patients provided a stool symptoms due to invasive Listeria monocytogenes specimen for the tests needed to determine the exact infections may not develop for 2-6 weeks (CDC, cause of their illness (Frenzen et al., 1999). A medical 1996). Chronic complications may take even longer to diagnosis can strengthen suspected linkages between a develop, often several weeks or months after the con- particular illness and food product, information impor- sumption of contaminated food. tant in pursuing foodborne illness litigation. The long incubation period for some pathogens makes it difficult to link specific contaminated food products Economic Costs of Foodborne Illness with adverse health outcomes, particularly the often nonspecific symptoms of foodborne illness. This time Because most foodborne illnesses are mild and do not lag also makes it difficult to trace problems back to require medical care, the average economic cost for ill specific producers, thus providing the food industry consumers and their families is likely to be small. with some protection from litigation. Several meals However, more severe illnesses can impose high mon- may have been eaten before people noticed foodborne etary costs, including medical costs and income or pro- illness symptoms. In general, the difficulty of tracing ductivity losses, as well as nonmonetary costs such as a foodborne illness back to a specific food source pain and suffering. increases as the incubation period lengthens. The Some economic costs of illness are shifted to parties longer the incubation period, the greater the number of other than the person who became ill. Types of cost other potential sources of foodborne illness and the shifting include: (1) insured medical expenses for lower the likelihood of having any samples of suspect those with health insurance are shifted to private or food available for microbiological testing. public health insurers; (2) uninsured medical expenses Mass Outbreaks of Illness. Public health authorities for those unable to pay their medical bills are absorbed are more likely to investigate mass outbreaks than by health care providers (or by taxpayers) when med- individual cases of foodborne illness. Therefore, out- ical expenses are deducted as a business loss; (3) time break cases tend to have more documentation, which lost from work by employees with sick leave benefits, can help determine liability for injuries due to food and reduced productivity by employees who report to products contaminated by microbial pathogens. work while ill, become costs for employers; and (4) Additionally, plaintiffs in outbreaks have the added medical expenses for foodborne illness covered under option of pursuing litigation via class action suits or government health plans (e.g., Medicare) are borne by mass litigation (see appendix). taxpayers. This cost-shifting reduces out-of-pocket costs for ill people and their families. The widespread Variations in Susceptibility to Foodborne Infections. prevalence of these cost-shifting mechanisms may People vary in their susceptibility to foodborne infec- reduce the economic incentives for ill individuals to tions due to host factors, such as age, stress, health of seek compensation from those responsible for causing their immune system, and personal hygiene, as well as their illness.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 7 Chapter 3

The Intersection Between Liability Law and Economics and Its Relevance to Foodborne Illness Litigation

Overview of the Incentive System of fresh juice, and prompted many juice manufacturers To Provide Food Safety to voluntarily begin pasteurizing juice products not previously pasteurized. The increasing number of U.S. firms that make or distribute food products have a foodborne illness outbreaks due to E. coli O157:H7 variety of incentives to reduce microbial pathogen con- contamination of unpasteurized juice products also led tamination to safe levels. A system of market, regula- the U.S. Food and Drug Administration (FDA) to pro- tory, and legal components provides these incentives to pose new regulations for juice products (Buzby and firms to produce safe food products (Garber, 1998a Crutchfield, 1999). These changes in market forces 1998b, extended to food safety). These incentives gen- and government regulations were in addition to the erally take the form of negative incentives or adverse adverse consequences for Odwalla, which included a consequences for firms responsible for selling voluntary product recall costing $12.5 million, a 17- pathogen-contaminated food. The basic components percent drop in revenue during the first 6 months after of this incentive system are: the outbreak, a record $1.5 million Federal fine for interstate shipment of an adulterated food product, and 1. Market forces: firms risk losing business reputa- 21 personal injury lawsuits (Buzby and Crutchfield, tion, market share, and sales revenue if consumers 1999; Roach, 1999; Munarriz, 1997). become concerned about safety problems with a firms products. Intersection Between 2. Food safety laws and regulations: firms that violate Federal, State, or local food safety laws or regula- Law and Economics tions may be subject to various penalties imposed by Of the three components of the food safety incentive courts or government agencies, including fines, prod- system, this report focuses on product liability and, uct recalls, and temporary or permanent plant closures. specifically, on how foodborne illness is treated under 3. Product liability law: firms found responsible product liability law. Product liability law specifies under product liability law for contaminated food the exact circumstances under which firms are held products that made people ill may have to pay finan- liable for injuries or deaths due to contaminated food cial compensation to the plaintiffs as well as punitive products, shifting some economic costs of foodborne damages. Firms also pay court costs and legal fees, illness from consumers to the firms responsible for regardless of most outcomes. causing illness. (See appendix for more information on the applicability of product liability law to food- Product liability law deals with products that are defec- borne illness.) For an overview about some of the lit- tive either because they pose hazards or because they erature on the intersection between law and economics are of inferior condition or quality (Keeton et al., 1984, and on how lawyers and economists view liability law p. 677). In essence, product liability operates within from different perspectives, see box, Intersection of this system of market, regulatory, and legal compo- Law and Economics (p. 10). This chapter outlines the nents and the combined incentives from this system role of insurance for foodborne illness because insur- encourage firms to provide safer products. Here we ance alters how the legal system provides incentives to are concerned only with products that pose microbial firms to produce safer food. The chapter concludes by food safety hazards. arguing that the combination of high transaction and The complexity of our food safety system and the information costs may lead to less than desirable levels interconnectedness of the three components is shown of food safety in the United States. by the 1996 outbreak of foodborne illness due to E. In a world of perfect information and competition, coli O157:H7 contamination of unpasteurized apple markets would penalize firms that produce unsafe juice manufactured by Odwalla, Inc. This outbreak products. Firms would receive negative signals about raised consumer concerns nationwide about the safety their errors and the market would correct itself.

8 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA Although the U.S. food market cannot be characterized State penalties for violation of regulations. The calcu- as having perfect competition, both economists and lation of these expected costs involves the probability legal analysts agree that liability law provides some of paying damages (i.e., compensation) in a lawsuit, economic incentives to avoid actions that may violate the size of the potential litigation costs, and any nega- laws. However, the extent to which liability law is tive impacts to the firms reputation and sales. Even if economically efficient is unclear. firms are not directly affected by legal action, many buy product liability insurance coverage to limit direct Economists tend to analyze liability law in terms of a losses due to lawsuits, transforming the expected risk search for efficiency in risk-bearing and incentives (see of lawsuits into a routine business expense. Cooter, 1991).6 Economic theory suggests that foodborne illness litigation provides signals to firms to The legal system could provide optimal deterrence if invest more in food safety, ultimately resulting in a firms could correctly anticipate the compensation that lower incidence of foodborne illness and an increase in the legal system would impose (Viscusi, 1989). general social welfare.7 One underlying economic However, three obstacles limit a firms understanding premise of foodborne illness litigation is that a firm of the true compensation they may have to pay follow- receives the appropriate incentives for efficient behav- ing a risky activity (Viscusi, 1989, p. 82). First, com- ior (e.g., to produce safer products) if it compensates pensation may not reflect the true damages that a all victims who are made ill by microbial contamina- plaintiff suffered. Second, compensation awards do tion of the firms food and if, for each case, the level not reliably indicate damages to actual victims. of monetary compensation equals total damages.8 In Victims who did not receive compensation, particularly reality, perfect compensation does not occur for all those who never filed a claim, are overlooked. Third, foodborne illness cases and outbreaks because relative- some firms do not contemplate paying compensation ly few firms are held responsible for monetary dam- awards, such as firms with short time horizons like ages caused by contamination of their food products. speculative ventures. Other firms, when faced with the possibility of particularly large claims, may limit lia- Potential liability is one part of firms anticipated costs bility by reorganizing under Federal bankruptcy law of operation, and firms will take the optimal amount of (Viscusi, 1989). food safety precautions consistent with minimizing their total costs of production (Johnson et al., 1989). Most food firms are insured, and insurers pay the loss- Doing everything physically possible to make food es and costs of any litigation and damages, so immedi- safe is not economically efficient. Instead, a firm will ate legal incentives for firms to produce safer food are incur increased costs of food safety precautions up to limited to the value of lost business reputation. the point where the marginal costs of these actions However, other legal incentives are likely to include equal the marginal expected benefits of reducing their future increases in insurance costs if insurers raise risks of being sued for a foodborne illness incident, rates or drop coverage for firms that cause insurance paying legal compensation, and paying any Federal or losses.

6 Economic efficiency occurs when a firms money and Insurance and Foodborne Illness other resources are allocated in such a way that additional Consumers ill with foodborne illness have reduced increases in an activity cannot be achieved without giving incentives to pursue legal claims if health insurance or up a portion of some other activity. employee benefits programs cover some of the costs of 7 Liability law deters or discourages food firms from operat- ing in ways that might result in foodborne illness: if firms their illness. Levit and Freeland (1988) found that, perceive a risk of incurring liability-related economic costs, after private and public insurance (e.g., social insur- they will allocate resources to reduce the probability that ance), patients directly paid for only 56 percent of all their products will cause foodborne illness. national health-care expenditures. They also found 8 Perfect compensation is a monetary compensation that that, in general, the more expensive the service, the restores a victim to the same level of well-being that would less (proportionally) did individuals pay (e.g., individ- have been enjoyed had the externality (e.g., illness or acci- uals directly paid only 10 percent of hospital-care dent) not occurred (Cooter, 1991). Of course, we cannot expenses). Because consumers tend to be risk averse, interpret this literally, as few, if any, plaintiffs would will- they want insurance coverage for when they become ingly acccept some externalities (Cooter, 1991), such as the ill. death of a child, paralysis, or chronic complications, in exchange for money.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 9 Firms also tend to be risk averse, and consequently paid and when it will be paid [in the case of they want insurance coverage to manage risks settlements]. (Shapiro, 1991). As part of a firms risk-management Comprehensive information about product liability strategy, firms purchase insurance to limit their expo- insurance coverage in the food industry is not readily sure to financial, legal, and other risks by sharing these accessible because the insurance industry is highly risks with insurance companies. Almost all defendant competitive and data about premiums and paid claims firms have third-party insurance, or liability insur- are valuable market information. One example of the ance for legal risks, and Clark (2000)9 neatly summa- insurance available to food firms is the products cont- rizes how insurance works for firms: amination coverage sold by the insurance subsidiary Large corporations typically have layers of of the National Food Processors Association. This excess insurance running into the tens or coverage includes assistance to deal with regulatory hundreds of millions of dollars. These excess investigations and media inquiries, as well as product insurers, in turn, insure themselves against the testing and compensation for the costs of product risk that they have to pay all of their money recalls, lost profits, and damage to brand names. out by going to the reinsurance market where Many food firms might obtain less comprehensive they acquire the ability to recoup the vast coverage. Further research is needed to establish how majority of their insured losses from reinsur- product liability coverage varies among different kinds ers. When a company is sued, its insurance of food firms. Some observers believe that nearly all company (a) provides a legal defense at the food firms have at least some coverage against food- insurers cost and (b) pays any resultant settle- borne illness due to a firms products (Clark, 2000). ment or judgment. Insurers, not defendants, In short, although both health-care insurance and lia- control litigation and determine what will be bility insurance benefit foodborne illness victims with compensation, they both distort incentives for firms to 9 Bruce T. Clark is an attorney at Marler-Clark, a law firm produce safer food. Health insurers seldom try to in Seattle that has handled many prominent foodborne ill- recover the costs of medical care for foodborne illness ness cases. patients from the firms responsible for contaminated

Intersection of Law and Economics

Adam Smith and Jeremy Bentham were among the first scholars to analyze the intersection of law and economics, and in the 1920's and 1930's, members of the legal-realist movement argued that the only function of tort law was to provide social insurance, since they believed tort law had no effect on the accident level (Landes and Posner, 1987). Although another wave of scholarship in law and economics occurred in the 1960's with seminal articles by Ronald Coase (common law as a mechanism for internal- izing social costs) and by Guido Calabresi (model of efficient accident law), sustained scholarship in this area lagged for another decade (Landes and Posner, 1987). Since that time, Richard Posner, Steven Shavell, Kip Viscusi, and others have added to the literature (e.g., Posner, 1997; Shavell, 1987; Viscusi, 1996). It is now widely accepted that law provides economic incentives to make safer products, though more research is needed on the extent and relative strength of these incentives, particularly compared with incentives from government regulation and market forces. A review of the legal and economic literature in the past decade suggests that lawyers and economists tend to emphasize different aspects of liability law. The legal literature commonly emphasizes that the objectives of liability law are to deter injurers and to compensate victims (e.g., Rose-Ackerman, 1991). The economic literature generally adds the concept of efficiency and analyzes law as a mechanism to achieve greater efficiency in risk management and incentives. This distinction is relevant here because this report not only investigates how the legal system deters production of unsafe food and compensates consumers for their foodborne illnesses, but also comments on efficiency and effectiveness issues.

10 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA food products because it is not cost effective. the lawsuit (Clark, 2000).10 Therefore, high informa- Consequently, health insurance limits the extent to tion costs and the uncertain outcome of a lawsuit may which food firms receive signals to produce safer food. provide greater disincentives for these plaintiffs to pur- Meanwhile, liability insurance distorts incentives to sue litigation than do transaction costs. Additionally, produce safer foods, particularly if insurers are most foodborne illnesses are relatively mild and short- involved in settlements that are kept confidential. lived and do not incur medical and other costs high enough to make litigation worthwhile for a plaintiff to pursue litigation. Similarly, for particular foodborne- Transaction and Information Costs illness cases, the potentially modest rewards may be Viscusi (1989) has argued that tort liability causes an too low to attract an attorney to represent a plaintiffs underproduction of health and safety in the United claim on a contingent fee basis. States because of high transaction and information However, plaintiff attorneys are not obligated to take a costs. The same claim can be made for foodborne ill- case on a contingent fee basis but may rather prefer a ness. High transaction costs and information costs dis- fixed fee or an hourly rate. In these situations, a plain- suade food-poisoning victims from filing lawsuits, hin- tiff must decide whether he or she can afford attor- der a plaintiffs ability to prove causality, limit feed- neys fees and costs to pursue a case. back to firms to produce safer food, and pose costs to plaintiffs and defendant firms. The result is a level of For defendant firms, a lawsuits transaction costs food safety that is less than the socially optimal level include lawyer fees, witness fees, time lost from usual provided by a perfectly competitive market. business activities (due to discovery requests for firm records, rehearsal of employee witnesses, courtroom appearances, etc.), potential loss of business reputation Transaction Costs associated with adverse publicity from a public trial Both plaintiffs and defendants may incur high transac- and consequent decline in sales, and increases in prod- tion costs (RAND, 1998a), that is, the amount of uct liability insurance premiums. The appendix dis- money spent for legal fees. Trials tend to cost any- cusses incentives for firms to settle rather than go to where from $50,000 to tens of millions (Kumamoto trial. and Henley, 1996). One study found that, of total expenditures on tort cases during 1983-88, 22 percent Information Costs were plaintiffs legal costs, 28 percent were defendants legal costs, and only 50 percent represented net High information costs may also discourage or prevent compensation received by plaintiffs (RAND, 1998a). potential plaintiffs from filing foodborne illness Large transaction costs suggest that liability is an claims. In particular, as discussed in the previous expensive way to compensate victims (Shapiro, 1991, chapter, plaintiffs may lack information to link an ill- p. 4). ness to their consumption of a specific food that was contaminated with a specific pathogen because of a For plaintiffs, transaction costs are primarily the costs specific defendants act or failure to act. And even if a of a product liability lawsuit (i.e., dollars spent for plaintiff could afford to spend a great deal of money lawyer fees, court filing fees, expert witness fees). for supporting information for a case, the information This definition can be expanded to include other costs, necessary for successful litigation may be unavailable. such as emotional stress, and money spent (e.g., travel costs) and time lost from work and other activities by The most crucial issue in litigation for foodborne ill- the plaintiffs and their families while preparing for and ness is causation, and epidemiologists and other med- appearing in court. ical professionals are generally needed to make such links. Plaintiffs typically try to obtain and use such Plaintiffs routinely pay attorneys at least a third of any expert medical or epidemiological testimony to support court award or settlement as a contingent fee (Cooter, their claim. For example, where possible, plaintiffs 1991). If plaintiffs pursue litigation under a contingent obtain information from local and State public health fee basis, then transaction costs are limited to time lost authorities as well as from the CDC, particularly for from work and other disruptions of daily activities, because their attorney assumes the responsibility for the financial costs associated with filing and pursuing 10 Of course, cost sharing depends on a fee agreement.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 11 outbreaks in which health authorities determined a are identified, the next step in the investigation is gen- food-pathogen linkage. Public health authorities who erally to identify the likely party that contaminated the investigate a case or an outbreak may be called to give food. Knowledge of the actual food-handling steps a deposition or to testify as to their findings in court. used by the implicated firm may be used. In outbreak However, because of conflicts of interest, they provide situations, circumstantial evidence may also be used, this information as part of their job rather than as an that is, information about the circumstances that help expert witness for the plaintiff (Rosenbaum, 2000). establish connections between a plaintiffs illness and Public health officials from outside an area and the the implicated food and firm. For example, evidence arena of the investigation may be used as experts that other people got sick after eating the same food (Rosenbaum, 2000). can lay the foundation for proving that the food was unwholesome (Draper, 1994). One common starting point in investigating causation is identifying the foodborne illness either by medical Information about medical expenses and costs of lost testing procedures or by simpler, descriptive tech- productivity helps support a plaintiffs claim. niques, such as determining if the ill consumers symp- Additionally, plaintiffs may also need information toms are consistent with typical acute and chronic about the costs of safety measures that the risk-reduc- symptoms for the particular foodborne illnesses being ing party might have taken (Viscusi, 1989, p. 72). considered. Information about what the consumer ate Monetary costs to a potential plaintiff of obtaining prior to the illness is needed to identify a likely food information necessary to prove causation may be high, source. If a pathogen is identified, the time frame and such information may not be available at all. under investigation can be narrowed down to match Similarly, the time necessary to obtain this information the typical incubation period of an implicated food- may also be high or infeasible. In general, costs are borne illness. Multiple food sources that could have lower if an illness is part of an outbreak investigated caused an illness complicate the determination of cau- by public health authorities or widely reported by the sation. If an illness and a suspect food (e.g., raw eggs) media.

12 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA Chapter 4

Analysis of Jury Verdict Data for Foodborne Illness

One purpose of this report is to perform a preliminary However, at least some foodborne illness jury verdicts exploration of product liability law for foodborne ill- went unreported in other areas of country. Jury verdict ness and the incentives it provides firms to produce reporting firms also claim that the product liability jury safer food. To this end, the analysis described here trials included in their databases were selected without used U.S. jury verdict data on foodborne illness law- any bias in favor of plaintiff verdicts or large awards. suits for 1988-97. We analyzed the frequency and size Thus, the foodborne illness jury verdicts identified by of awards by illness severity, pathogen, and food sub- the firms in areas without complete reporting appeared categories. We used multivariate analyses to examine to be a representative sample of all foodborne illness the simultaneous effects of the various factors that jury trials in these areas. might affect whether or not plaintiffs receive food- We searched the Westlaw and Lexis-Nexis (WLN) borne illness jury verdicts and the size of the award. databases for every jury verdict involving personal injuries due to pathogen-contaminated food between U.S. Foodborne Illness Court Data 1988 and 1997. Relevant jury verdicts in the WLN databases were identified by searching for the general The universe of U.S. foodborne illness cases is classification term food poisoning or the names of unknown because there is no national system docu- common foodborne pathogens or illnesses (botulism, menting all product liability cases, particularly those Campylobacter, campylobacteriosis, ciguatera, cigua- dropped or settled out of court. Therefore, we focused toxin, Clostridium, Cryptosporidium, Cyclospora, E. exclusively on foodborne illness lawsuits that were coli, hepatitis, Listeria, listeriosis, Salmonella, tried and resulted in jury verdicts. We identified food- Shigella, Staphyloccocus, Toxoplasma, toxoplasmosis, borne illness jury verdicts by systematically searching Trichinella, trichinosis, Vibrio, and Yersinia). two major jury verdict databases, the West Groups Westlaw Jury Verdicts and Settlement Summaries Cases were included only if they met our definition of (West Group, Inc., Eagan, Minnesota) and the Lexis- a foodborne illness case: the illness (1) produced Nexis Verdicts Library (Reed Elsevier PLC, London, symptoms consistent with foodborne illness (e.g., gas- England). Although these databases include both trials trointestinal distress) in the medical literature, (2) was and settlements, we discarded the settlement data as linked to food, and (3) was claimed to have resulted they were not representative.11 These databases from pathogens or foreign objects that are organic or included descriptive summaries of civil jury verdicts, could have been living in the food. gathered by jury verdict reporting firms that collect The data include lawsuits whose trial dates or dates of and sell information about legal cases for use by prac- resolution were between 1988 and 1997. The analysis ticing attorneys. Information provided about each ver- was limited to the 1988-97 period because most of the dict was limited, and omitted whether verdicts had jury verdict reporting firms did not begin providing been overturned or reduced through the post-trial their case summaries to Westlaw or to Lexis-Nexis appeal process. until the late 1980s. Most jury verdict reporting firms cover a single State The results from the WLN databases search were sup- or metropolitan area, although none collect every ver- plemented with information from a few published case dict in their covered area. In some areas of the country histories. We identified 178 foodborne illness jury with competing firms, the firms believed that every trials in 32 States that reached legal resolution during foodborne illness jury verdict in the area was reported the 10-year period, 1988-1997. None of the 178 jury by at least one firm, assuring complete reporting. trials were class action suits, although some involved more than one injured plaintiff, such as a family with the same foodborne illness. 11 For example, settlement data are prone to selection biases such as the differences between confidential settlements and After identifying the foodborne illness jury verdicts, settlements that are disclosed to the public. we then searched the entire Lexis-Nexis Litigation

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 13 Library database to determine whether each foodborne dants, cases where plaintiffs won but did not receive illness jury verdict had been appealed to a higher compensation, and cases for which we did not have court. Two lawsuits had been unsuccessfully appealed information on awards. by defendants. Several others had been remanded for Expected award: This is the mean plaintiff award retrial, but we were unable to determine the outcome multiplied by the percent of foodborne illness jury tri- so they were dropped from the analysis. als won by plaintiffs. The only cases excluded here We compared the verdicts from the areas with full are those for which we did not have information on reporting and sample reporting to determine whether awards. This is the most relevant statistic for the pur- there were any systematic differences that might reflect pose of this report as it shows the expected monetary sampling bias in the areas with sample reporting. The transaction as a result of a foodborne illness lawsuit comparison was performed by conducting a likelihood between different parties in the chain of food produc- ratio test of multivariate models predicting the out- tion, distribution, and consumption. come of trials in each area. The test revealed no significant difference between the population of verdicts drawn from each area (c2=13.9, d.f.=10, P>.05), sug- Findings gesting that the sampling was unbiased. Based on this evidence, the 178 cases identified by searching the Time to Legal Resolution WLN databases appear to provide representative infor- In general, the time to a legal resolution is the sum of mation about foodborne illness lawsuits resulting in several components: jury verdicts in areas with both full and partial jury verdict reporting during 1988-97. l the time between the incident (e.g., illness or injury) and the filing of the complaint, A number of different characteristics of each foodborne illness lawsuit were coded in a computer data- l the time between the filing of the complaint and the base, including the date of the incident that caused the beginning of the trial, injury, the type of food and pathogen involved, the the time between the beginning of the trial and the severity of the illness, the date of the trial verdict, the l announcement of the verdict, outcome of the trial, and the amount of any damage award. Descriptive information in the published WLN l the time between the verdict and the filing of an summaries varied. For example, all cases reported the appeal (if one is filed), outcome/verdict, but three cases did not report compensation amounts, and few reported claimed expenses. l the time it takes for the appeal to a final judgment (if an appeal is filed), and Financial damage awards were updated to 1998 dollars using the Bureau of Labor Statistics annual Consumer l the time between a conclusive plaintiff victory and Price Index for all urban consumers to ensure compa- the receipt of any compensation. rability between lawsuits.12 For many of the subse- Of the 178 court decisions for foodborne illness law- quent sections, we report three main statistics about suits in the WLN data, 110 observations provided awards: information on the time between the alleged foodborne Mean award: This is the average monetary compensa- illness incident and the legal resolution. The average tion awarded to plaintiffs who received compensation. time elapsed between the date of the incident that This statistic excludes cases won by defendants, cases resulted in illness and the date of the jury verdict was where plaintiffs won but did not receive compensation, 3.1 years (median of 2.8 years). One case was not and cases for which we did not have information on tried for nearly 10 years, although another case was awards. tried in just 5 months. The WLN data do not permit a more descriptive breakdown of time components. Median award: This is the midpoint of monetary compensation awarded to plaintiffs who received com- The average time required to bring a foodborne illness pensation. This statistic excludes cases won by defen- lawsuit to trial was comparable to the average time elapsed between the filing of a product liability lawsuit and the trial verdict (2.5 years) reported in a study of all product liability cases in five States during 1983-85 12 Source: . [Accessed Feb. (GAO, 1989, p. 49). The slow pace of foodborne ill- 22, 1999]. ness litigation may impose significant costs on con-

14 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA sumer plaintiffs. Nevertheless, delay may be advanta- receive this amount if they went to trial, less their legal geous for consumers who developed long-term chronic and court fees which typically total about one-third of complications of foodborne illness and who were well the award. Conversely, firms could expect to pay this advised to wait to determine the full extent of their amount if they go to trial, in addition to their legal fees injuries before filing a lawsuit specifying monetary and any other costs associated with a public trial such damages (Rosenbaum, 1998). as loss of business reputation.

Frequency and Size of Awards Court Awards by Severity Category Most plaintiffs failed to convince juries that defendants The jury verdict summaries provided only minimal were legally responsible for causing their illness. One- information on illness severity. Therefore, we subdi- third of the verdicts (31.4 percent) resulted in a mone- vided the 178 court cases into three categories by tary award for the consumer. For the 55 cases where severity: cases involving a premature death, cases the plaintiffs prevailed, the mean award to plaintiffs involving hospitalization but not a premature death, was $133,280 (table 2). However, the distribution of and all other cases involving less severe illnesses. Six awards was highly skewed because some awards were lawsuits (3 percent) involved a death, and another one- much larger, including two awards over $1 million. third of the lawsuits (60 percent) involved nonfatal The median award of $25,560 consequently provides a injuries severe enough to require hospitalization. The better indication of the typical jury award for damages average length of hospitalization was 9 days, although resulting from foodborne illness. The total amount one plaintiff was hospitalized for 49 days. awarded to the 55 cases that prevailed in court was Injury severity is a major factor affecting an expected $7.3 million, but the two largest awards accounted for award. Of the six lawsuits involving a premature over half (51 percent) of this sum. death, juries awarded damages in four (66.7 percent). An alternative measure of the amount awarded in food- The expected award for a lawsuit that claimed a pre- borne illness lawsuits is the expected award. The mature death as a result of a foodborne illness was expected award is the mean award multiplied by the $183,053, far higher than the expected award percent of jury trials won by plaintiffs (i.e., 31.4 per- ($44,713) for lawsuits involving nonfatal foodborne cent). The expected award was $41,888, nearly two- illnesses requiring hospitalization, and the expected thirds larger than the median plaintiff award. award ($32,563) for all other illnesses (table 3). This Consumers and firms involved in foodborne illness pattern is consistent with the finding of the 1989 GAO lawsuits could take this expected award into account report on general product liability cases in five States when making decisions about whether to resolve a in which the size of the award varied by the type and lawsuit prior to trial. Consumers could expect to degree of injury.

Table 2—Compensation for consumer plaintiffs in foodborne illness lawsuits decided by jury verdicts, 1988-971

Outcome Sample Percent Range of Mean Median Expected Total size won by compensation award award award2 amount plaintiffs compensated

Number Percent ------1998 Dollars ------

Plaintiff3 55 100 2,256-2,368,858 133,280 25,560 133,280 7,330,412

Defendant4 120 0 0 0 0 0 0

Total 175 31.4 0-2,368,858 133,280 25,560 41,888 7,330,412

1 Data updated to 1998 dollars using Bureau of Labor Statistics Consumer Price Index for all urban consumers. Of the 178 court decisions, 175 had published information on awards. 2 The expected award is the average award multiplied by the percent of foodborne illness jury trials won by plaintiffs. 3 Plaintiff verdict or award combined. 4 Defendant verdict or judgment combined. Occasionally, unsuccessful plaintiffs covered defendants’ court costs but these were not enumerated here.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 15 Table 3—Compensation in foodborne illness court cases Table 4—Foodborne pathogens, toxins, or by severity category, 1988-971 illnesses involved in foodborne illness lawsuits decided by jury verdicts, 1988-97 Illness Court cases Percent Mean Median Expected severity with award won by award award award2 Pathogen Lawsuits information plaintiff Number Percent1 Number Percent ------1998 dollars ------Salmonella (any serotype) 39 21.9 Premature death 6 66.7 274,580 185,828 183,053 Hepatitis (any type) 10 5.6 Staphylococcus 6 3.4 Hospitalized Vibrio vulnificus 6 3.4 and survived 60 31.7 141,199 61,814 44,713 Shigella (any type) 5 2.8 Campylobacter 4 2.2 Other cases 109 29.4 110,916 11,746 32,563 Mold 4 2.2 E. coli 2 3 1.7 Total 175 31.4 133,280 25,560 41,888 Botulism (Clostridium botulinum) 2 1.1 Ciguatera 2 1.1 1 Only 175 of the 178 court decisions had award information so the Salmonella and Staphylococcus award totals do not represent statistics for all court awards. (combined) 1 0.6 2 The expected award is the mean plaintiff award multiplied by the Streptococcus 1 0.6 percent of foodborne illness jury trials won by plaintiffs. The only Trichinella spiralis 1 0.6 cases excluded here are cases for which we did not have information Vibrio parahaemolyticus 1 0.6 on awards. Adverse reaction to protective immunization after exposure to foodborne hepatitis 1 0.6

Court Awards by Implicated Pathogen Not specified 92 51.7 The ability of consumer plaintiffs to identify the spe- Total 178 100 cific pathogen and food item that made them ill is like- 1 Percents may not add to 100 due to rounding. ly to have an important effect on the outcome of a trial 2 The case summaries for the three lawsuits involv- because of the emphasis placed on establishing a ing E. coli did not mention the serotype, but all three cases appeared to involve E. coli O157:H7. causal link between a defective product and the alleged injury under product liability law. Less than half (48 Court Awards by Implicated Food percent) of the jury verdict reports implicated a specific foodborne pathogen, toxin, or illness (table 4). Most jury verdict reports (92 percent) identified some Some reports may have failed to record pathogen kind of food as the cause of illness (table 6). names, so the actual proportion of lawsuits that impli- However, one-fourth of the reports simply named cated a specific pathogen might be somewhat higher. meals such as dinner or food categories such as fast Among the jury verdict reports that named a pathogen, food that presumably included multiple food items, Salmonella was the most frequently cited pathogen, leaving the precise source of illness unclear. In con- followed by hepatitis (any type).13 trast, two-thirds of the jury verdict reports (66 percent) Plaintiffs who alleged illness from a specific pathogen identified a specific food item or food as the cause of were more likely to receive compensation (41.7 per- illness. The most frequently mentioned foods were cent) than plaintiffs who did not implicate a specific sandwiches, followed by seafood (excluding oysters) pathogen (22 percent), and the expected award was far and chicken. Only three lawsuits mentioned packaged higher when a specific illness or pathogen was alleged. meals such as canned foods or frozen meals, suggest- These findings suggest the importance of establishing ing that litigation involving packaged meals was either a causal link for a plaintiff to prevail in a foodborne uncommon or likely to be resolved out of court. illness trial and receive compensation (table 5). Interestingly, cases whose jury verdict summaries alleged a specific food as the cause of the illness resulted in a lower percentage of favorable plaintiff verdicts (26.3 percent) than cases that did not name a 13 Knowledge about the different pathogens varies greatly, specific food (e.g., dinner or fast food) (41.0 per- meaning that there is uneven documentation, scientific literature, and legal precedent for the different pathogens in cent) (table 7). This finding is counterintuitive foodborne illness litigation. because of the importance of establishing a causal link.

16 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA Table 5—Compensation in foodborne illness court cases by pathogen category, 1988-971

Pathogen Court cases Decision Mean Median Expected category with award for award award award2 information plaintiffs

Number Percent ------1998 dollars ------

Alleged illness from a specific pathogen 84 41.7 197,599 55,061 82,333

Unspecified pathogen 91 22.0 20,722 11,960 4,554

Total 175 31.4 133,280 25,560 41,888

1 Of the 178 court decisions, 175 had award information. Therefore, the award totals do not represent statistics for all court awards. 2 The expected award is the mean plaintiff award multiplied by the percent won by plaintiffs. The only cases excluded here are cases for which we did not have information about awards.

This finding may reflect a lack of detail about the trials reported in the jury verdict summaries. Some trials may have alleged a specific food, but jury verdict Table 6—Food items involved in foodborne illness lawsuits decided by jury verdicts, 1988-97 reporters did not consider this information important to record. However, the expected award was higher for Food item Lawsuits those alleging a specific food ($48,593) than for those 1 who did not ($29,358). Number Percent Single vehicle Sandwiches (excluding hamburgers Court Awards by Type of Defendant and egg sandwiches) 15 8.4 Seafood (excluding oysters) 11 6.2 Plaintiffs may sue multiple defendants for several rea- Chicken 10 5.6 sons. For example, even in cases where there is a Hamburgers and ground beef 9 5.1 Oysters 9 5.1 strong indication of wrongdoing by restaurants, such as Salad 7 3.9 documented improper cooking temperatures, plaintiffs Sausages and unknown meat 5 2.8 may think that the illness was first caused further back Beverages (excluding milk) 5 2.8 in the food production chain by the pathogen contami- Mexican food 5 2.8 Baked goods (excluding nating the product or by sloppy slaughtering practices desserts with raw egg) 4 2.2 or poor sanitation in processing, and they wish to hold Chinese food 4 2.2 all parties accountable (Rosenbaum, 2000). Alterna- Packaged meals (e.g., canned food, TV dinner) 3 1.7 tively, suing multiple defendants may also be a sign Pork 3 1.7 that the plaintiff does not have sufficient evidence of Ice cream 2 1.1 causation to isolate and name one defendant (Clark, Beef (excluding hamburgers and 2000). ground beef) 2 1.1 All other single vehicle (e.g., honey, Table 8 reports foodborne illness lawsuits by defendant lasagna) 19 10.8 type. Of the 178 court cases, 135 (75.8 percent) named one defendant, 30 (16.9 percent) named two Multiple vehicle (e.g., “restaurant food,” defendants, and 13 (7.3 percent) named three or more “fast food,” “dinner”) 46 25.8 defendants, for a total of 234 separate defendants.14 Not specified 15 8.4 Most defendants were restaurant franchises with parent companies, the second largest category of defendants. Total 178 100 1 Percents may not add to 100 due to rounding.

14 We did tabulations on up to only three defendants per case because almost all cases had three or fewer defendants. The total number of defendants is slightly underestimated because a few cases had four or more defendants.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 17 Table 7—Compensation in foodborne illness court cases Table 8—Defendants in foodborne illness by food category, 1988-971 court cases by firm type, 1988-97 1 Food Court cases Decision Mean Median Expected Defendant Total defendants category with award for award award award2 information plaintiff Number Percent

Number Percent ------1998 dollars ------Restaurants 74 31.6 Foodstores 27 11.5 Alleged illness from Distributors 11 4.7 2 a specific food 114 26.3 184,652 27,584 48,593 Manufacturer 29 12.4

Did not specify food 61 41.0 71,634 18,707 29,358 Parent 60 25.6 Other3 33 14.1 Total 175 31.4 133,280 25,560 41,888 Total 234 100.0 1 Only 175 of the 178 court decisions had award information so the 1 award totals do not represent statistics for all court awards. Of the 178 court cases, 43 had multiple defen- 2 The expected award is the mean plaintiff award multiplied by the dants for an overall total of 234 defendants. percent won by plaintiffs. The only cases excluded here are cases Tabulations were performed on up to three defen- for which we did not have information on awards. dants per case. The number of defendants is underestimated for cases with more than three defendants because of insufficient information. 2 Other Information Provided Includes producers such as dairies and egg farms. 3 Includes nine people, five hotel restaurants, three by Court Data insurers, two each of clubs, cruise lines, food service, and vending machine firms, and 1 each of casi- Of the 175 court cases with award information, public no, deli, department store, fair vendor, government health authorities were involved in 23 (13 percent), entity, investor, psychiatric institution, and railroad. and of these 23 lawsuits, plaintiffs won 11 (47.8 percent). Both plaintiffs and defendants used expert wit- from pursuing litigation (e.g., court costs and legal nesses. Expert witnesses, such as physicians, are like- fees). Although the WLN data provided some infor- ly to be called only when their testimony is considered mation about the costs of illnessmedical expenses essential. Plaintiffs called one or more physicians as and lost productivity, for instancethe data did not expert witnesses in 67 percent of the foodborne illness provide information about the costs of pursuing litiga- lawsuits. In contrast, only 45 percent of the defendants tion, and the data were too weak to compare the costs called physicians. The disparity in the use of medical of illness for each case against the monetary outcome experts shows that establishing the role of a foodborne of each case. pathogen in causing an illness was a more important issue for plaintiffs because they have the burden of The data do, however, provide some insight into the proving that a food caused the illness. magnitude of the claimed medical expenses and claimed lost productivity. Of the 178 court cases, 81 provided at least partial information on claimed med- Key Findings ical expenses, ranging from $18 to $342,830, with a median of $5,612 (mean $19,292), and 42 provided Despite their greater reliance on medical experts, most some information about claimed lost productivity, consumer plaintiffs failed to convince juries that defen- ranging from $30 to $274,966, with a median of dant firms were legally responsible for causing their $1,905 (mean of $20,151) (all in 1998 dollars). illness. One-third of the foodborne illness lawsuits (31 percent) resulted in a monetary award for the con- Even with a favorable verdict, plaintiffs may not sumer. Only a few of the jury verdict summaries pro- receive damages set by a jury if: (1) the defendant vided commentaries describing why juries decided in does not have enough money or insurance to cover the favor of firms rather than consumers. Some of the award, (2) the award exceeds a State-mandated cap, or specific reasons cited for deciding in favor of firms (3) a jury award is reduced by the trial judge or during included failures by plaintiffs to prove that a food settlement discussions prior to appeal, or on appeal product was defective or to prove that the plaintiff (Broder, 1986). actually consumed the food product. Information about the defenses used for the foodborne However, even if plaintiffs receive awards, the awards illness lawsuits is incomplete and may not be represen- may not cover the costs incurred from the illness and tative of all defenses. However, in 26 lawsuits, the

18 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA defense argued that no one else became ill with food- A second lawsuit characteristic is whether the plaintiff borne illness. In 10 lawsuits, the defense argued that can provide sufficient evidence linking his/her illness the plaintiff had a pre-existing illness that either was to a specific foodborne pathogen that may have been in the cause of the current illness or was the current ill- a food produced by the defendant. This information is ness. In 19 lawsuits, the defense argued that the tim- likely to have an important effect on the outcome ing of the specific foodborne illness after ingestion of because of the emphasis that the law places on estab- allegedly contaminated food was inconsistent with the lishing a causal link between an illness and a product. incubation period of that foodborne illness. In six law- A third lawsuit characteristic is whether a public health suits, the defense admitted liability. authority was involved. This characteristic is important because the critical issue in most litigated food- Multivariate Analyses borne illness lawsuits is causation: whether the plain- of the Court Data tiff can prove that his/her illness resulted from exposure to the particular food item at issue. In outbreak We consider three general factors potentially affecting situations, plaintiffs often rely upon the investigating foodborne illness jury verdicts: lawsuit characteristics, public health authority to supply the epidemiological plaintiff characteristics, and defendant characteristics. link; in cases where the public health authority cannot establish the link, the plaintiffs case may be weaker. Lawsuit Characteristics A fourth lawsuit characteristic is whether plaintiffs or defendants used medical witnesses to support their In addition to the strength of the case, plaintiff litiga- case. However, expert witnesses, such as physicians tion success rates and the amount of awards in court who receive large consulting fees, are likely to be trials generally vary by jurisdiction (e.g., small claims, called only when their testimony is considered essen- civil, State, Federal, county) and by the nature of the tial. Merritt and Barry (1999) found that in product injury, which is correlated with the type of law (e.g., liability lawsuits, plaintiffs were more likely than contract, property, torts, and their subcategories such defendants to employ expert witnesses. This makes as product liability in the case of tort law) (Eisenberg, sense as plaintiffs have the burden of proof. Merritt 1991; Daniels and Martin, 1986). Geographic varia- and Barry did not explore the impact of using these tions may arise because of differences in propensity to witnesses on case outcomes or awards. sue, access to lawyers and the legal system, or State laws. Plaintiff Characteristics Therefore, the first characteristic of a lawsuit considered here is the State where it is filed. Because data Plaintiff characteristics might influence foodborne ill- were not available by State on all outcomes of the ness trial outcomes. Children and the elderly are cate- foodborne illness subset of personal injury lawsuits or gories of people particularly at risk from the more for personal injury lawsuits as a whole, available severe complications of foodborne illness. Merritt and Federal data were used to better understand regional Barry (1999) found tentative indications that minors differences. Of all Federal district-court civil cases, were more likely than adults to win malpractice there were 1,898 completed trials with reported out- claims. While jurors are not supposed to act on sym- comes during 1988-97 for the personal injury subcate- pathy, they are inclined to favor the plaintiff in cases gory of product liability (Eisenberg and Clermont, involving children (Clark, 2000). 2000). After aggregating by State, plaintiff success rates ranged widely from zero to 66.7 percent, while Illness severity is another plaintiff characteristic likely mean awards ranged widely from $0 to $8,160,156.15 to be a factor in whether or not a plaintiff prevails, per- If these results hold for the foodborne illness subset of haps partly because more severe cases (e.g., hospital- personal injuries, plaintiffs in some areas of the coun- ized cases) tend to have more testing and better docu- try may be more likely to win foodborne illness trials mentation to support plaintiffs claims. Merritt and and receive larger awards than plaintiffs in other areas. Barry (1999) used a 12-point scale to rate various degrees of injury to the plaintiff ranging from 1 (pure property damage and no injury to health) to 12 (death), and found that the most severely injured plain- 15 Updated to 1998 dollars using the Bureau of Labor tiffs were not more likely to win in court. This out- Statistics consumer price index (CPI) for all urban con- come may result because severe cases tend to involve sumers.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 19 larger damages, in turn increasing the defendants variable, which indicated whether the plaintiff pre- insurers incentive to contest the lawsuit. Other stud- vailed. Ordinary least squares (OLS) regression was ies indicated that awards varied by injury type and used for the Award Model. The dependent variable for severity (GAO, 1989; Rodgers, 1993) and were higher the award amount was highly skewed, so it was trans- in tort trials involving death (Tabarrok and Helland, formed to a logged variable.16 1999). In addition to medical costs and lost productivity, some Independent Variables plaintiffs in foodborne illness lawsuits claimed other and Their Hypothesized Effects damages, such as emotional distress, loss of consortium (i.e., a spouses help and affection), and pain and The Win Model suffering. A dollar value for these damages is difficult Table 9 presents the definitions and mean values for to assess because of their subjective nature. If plain- the independent variables used in both models. The tiffs can provide convincing evidence that these com- Win Model included 11 independent variables, which plications existed and can be valued fairly, then awards are hypothesized to affect the odds of an award. These might be higher. independent variables can be categorized into lawsuit, plaintiff, and defendant characteristics. Defendant Characteristics As a proxy for the unmeasured regional differences potentially affecting the outcome of a lawsuit, a vari- Defendant characteristics may help explain foodborne able representing plaintiff success rates for personal illness outcomes as well. According to one hypothesis, injury lawsuits by State (RATERAW) was included in if juries perceive that certain defendants can afford to the model. The data are from an Internet-accessible pay more (i.e., have deep pockets), they tend to make database of Federal district-court civil cases with these defendants pay higher awards than otherwise. reported outcomes, 1988-97 (Eisenberg and Clermont, Although some studies, such as Tabarrok and Helland 2000). States with higher plaintiff success rates in (1991), in the case of jury trials, and Shanley (1991), Federal trials for personal injury lawsuits are anticipat- in the case of awards paid to plaintiffs after post-trial ed to have higher plaintiff success rates in jury deci- award adjustments, appear to support this hypothesis, sions for the foodborne illness subcategory of personal Vidmar (1997) raised important questions about the injury lawsuits. validity of this hypothesis such as whether there were plausible alternative explanations. And, proponents of Four other independent variables measured the charac- the deep-pocket hypothesis did not investigate whether teristics of the lawsuit. It was anticipated that lawsuits affluence affects which side will prevail in court. implicating specific foodborne pathogens Plaintiff victories may be less likely in cases against (PATHOGEN) were more likely to result in plaintiff deep-pocket defendants since such defendants may victories and that public-health officials involvement have greater incentives to protect their corporate repu- (PUBLIC) would provide substantiating information tations and market share (i.e., more to lose) and have for a plaintiffs claim, increasing the chances of a greater financial resources with which to do so (e.g., plaintiff award. It was also anticipated that the plain- hire more and better experts and lawyers). Overall, tiffs chances of prevailing would be increased by however, the impact of deep pockets is still unclear, plaintiffs use of medical expert witnesses (PWIT- particularly if jurors tend to believe that all defendant DOC) and reduced by defendants use of these wit- food firms have insurance that might cover any award nesses (DWITDOC). they might decide to give the plaintiff. Another variable, YEAR1993, indicates whether a lawsuit was resolved in 1993 or later, in anticipation that Design of Analysis plaintiffs are more likely to prevail after 1993 because of increased public awareness of food safety hazards We performed two multivariate regression analyses to and related litigation after the large 1993 outbreak examine the effects of the various factors that might influence whether or not plaintiffs win foodborne illness jury verdicts (Win Model, n=175) and the size of the award in the case of plaintiff victories (Award 16 Sample selection models could have also been used here, Model, n=55). For the Win Model, the logit model but we decided that logit and OLS regressions were appro- was selected to handle the dichotomous dependent priate for the purpose of this report.

20 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA from undercooked hamburgers contaminated with E. Two variables measured defendant characteristics. The coli O157:H7. This outbreak was one of the most first variable (DEEPPOCK) was included to measure well-publicized outbreaks in terms of incidence, sever- the impact of defendants with deep pockets on the ity, and legal and economic ramifications, and could plaintiffs odds of winning. The criteria used to deter- have affected jurors perceptions of the role of firms in mine whether a defendant had deep pockets were pre- causing foodborne illness. sumed to be met if there was any evidence that the defendant had one or more of the following: (1) three The Win Model also includes two independent vari- or more retail operations, (2) 40 or more full-time ables (HOSPITAL, DEATH) measuring the severity of employees, (3) two or more manufacturing plants, or the plaintiffs illness. We hypothesize that more severe (4) three or more warehouses. Two other defendants, a illnesses are more likely to result in plaintiff victories State government unit and a country club, were also because of greater medical documentation and because considered to have deep pockets. Under these criteria, juries might be more sympathetic toward these cases. roughly half of the lawsuits in our sample had at least Similarly, a variable indicating whether or not one or one deep-pocket defendant. The anticipated relation- more of the plaintiff(s) was a child (CHILD) was ship for this variable is unclear. The second variable included in anticipation that juries may be more sym- indicated whether one or more defendants was a pathetic toward child or infant plaintiffs. restaurant (REST). We anticipated that the plaintiffs chances of winning would increase if plaintiffs and

Table 9—Definitions and mean values of independent variables

Win model Award model Variables (win/no win) (award size) (N=175) (N=55)

RATERAW (the weighted average win rate per State in completed Federal District product liability jury trials for personal injury) .3005

PATHOGEN (1 if a specific foodborne pathogen, toxin, or illness was implicated; 0 otherwise) .4800

PUBLIC (1 if a public health authority was involved; 0 otherwise) .1314

PWITDOC (1 if the plaintiff employed one or more doctors as expert witnesses; 0 otherwise) .9657

DWITDOC (1 if the defendant employed one or more doctors as expert witnesses; 0 otherwise) .5600

YEAR1993 (1 if the lawsuit was resolved in 1993 or later; 0 otherwise) .5086 .3818

HOSPITAL (1 if the plaintiff(s) was hospitalized; 0 otherwise) .3429 .3455

DEATH (1 if the lawsuit involved a death; 0 otherwise) .0343 .0727

CHILD (1 if one or more of the plaintiff(s) was a child; 0 otherwise) .0914 .1091

DEEPPOCK (1 if one or more of the defendants had “deep pockets”; 0 otherwise) .5429 .3455

REST (1 if one or more of the defendants was a restaurant; 0 otherwise) .4229

CHRONIC (1 if the lawsuit involved a chronic complication; 0 otherwise) .1636

DISTRESS (1 if plaintiffs claimed emotional distress; 0 otherwise) .1091

LOSSCONS (1 if plaintiffs claimed loss of consortium; 0 otherwise) .1091

PAINSUFF (1 if plaintiff claimed pain and suffering; 0 otherwise) .1455

AVGRAW (the average award in thousand dollars per State in completed Federal District product liability trials for personal injury) $2,081.78

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 21 Table 10—Multivariate analyses of foodborne defendants are relatively close together in the chain of illness jury verdicts food production, distribution, and consumption, such as the relationship between a restaurant and a restau- Win model Award model (win/no win) rant customer, facilitating the identification of linkages between an illness and a responsible firm. Predicted Logit Odds ratio1 Regression relationsihp coefficient coefficient The Award Model

INTERCEPT —— 0.77 —— 9.29 The Award Model has 10 independent variables, 5 of (0.94) (0.43) which were also included in the Win Model: HOSPI- TAL, CHILD, DEATH, YEAR1993, and DEEPPOCK. RATERAW + -2.29 0.10 (2.51) All five were anticipated to have positive relationships with award amount. An additional severity measure PATHOGEN + 0.97* 2.63 was added for chronic complications (CHRONIC) in (0.41) anticipation that the associated higher costs would PUBLIC + 0.28 1.32 increase awards. The Award Model also included vari- (0.58) ables for claims such as emotional distress (DIS- PWITDOC + 0.06 1.06 TRESS), loss of consortium (LOSSCONS), and pain (0.24) and suffering (PAINSUFF); all were expected to raise award amounts. The final variable is the average DWITDOC - -0.93** 0.39 Federal personal injury award (in thousand dollars) by (0.36) State (AVGRAW) using the Eisenberg and Clermont YEAR1993 + -0.64 0.53 -0.22 (2000) data. We anticipated that States with higher (0.39) (0.38) Federal awards for personal injury lawsuits would also HOSPITAL + -0.07 0.94 1.12** have higher jury awards for the foodborne illness sub- (0.43) (0.39) category of personal injury lawsuits.

DEATH + 1.34 3.82 3.06** (1.02) (0.67) Regression Results and Interpretation CHILD + 0.53 1.69 0.28 The Win Model (0.62) (0.60) Table 10 presents the findings of the multivariate DEEPPOCK ? -1.24** 0.29 0.41 (0.39) (0.38) analyses. The logit regression provides an odds ratio which measures the multiplicative effect of the inde- REST + -0.51 0.60 pendent variables on the odds of a plaintiff judgement. (0.40) The max-rescaled R2 is 0.2791 for this model (a pseu- CHRONIC + 1.62** do-R2 for models with binary dependent variables, pro- (0.47) vided by SAS).17 DISTRESS + 0.95 In the Win Model, DEEPPOCK, DWITDOC, and (0.60) PATHOGEN all had statistically significant effects on LOSSCONS + 0.82 the plaintiffs chances of winning. When the lawsuit (0.57) involved one or more deep-pocket defendants, the odds of a judgment for the plaintiff decreased by 29 percent, PAINSUFF + 0.25 (0.50) suggesting that defendants with greater financial resources were either more successful in their defense AVGRAW + -0.0001 or more likely to settle stronger cases out of court. (0.0001)

Sample size 175 55

*Significant at the .05 level. 17 For a general linear model, Cox and Snell (1989, pp. **Significant at the .01 level. 208-209) proposed a generalized coefficient of determina- Numbers in parentheses are standard errors. 1 tion, and this was later adjusted by Nagelkerke (1991) for The odds ratio is the marginal logit effect. use in models with binary dependent variables (i.e., the max-rescaled R2).

22 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA When the defendants used medical experts, the odds of Discussion a plaintiff victory decreased by 39 percent. This relationship was expected as a defendants medical experts Our findings emphasize the importance of consumer can highlight weaknesses in a plaintiffs case and place plaintiffs identifying the particular pathogen that doubt in the minds of jurors as to the integrity of any caused their illness, and whether or not the illness was linkages between a particular foodborne illness and the caused by consumer or firm error. The award model implicated food product. Implicating a specific food- showed that juries base monetary awards on illness borne illness or pathogen that caused the illness severity while relatively subjective considerations, increased the odds of a plaintiff victory by over 260 such as pain and suffering, emotional distress, and loss percent, suggesting that the stronger the link between of consortium, were not significant factors. Also, an ill individual and a particular foodborne pathogen, whether the plaintiff was a child or whether the defen- the stronger the plaintiffs case. dant had deep pockets were not significant factors. The Award Model Undoubtedly the greatest research gaps concern how often foodborne illness lawsuits are filed, how many After controlling for the fact that plaintiffs receive are settled or otherwise resolved before trial, and how larger awards for product liability injury trials in some settlements differ from jury verdicts. And, as settle- areas of the country than others, three variables in the ments comprise the bulk of all outcomes and have dif- Award Model were significant at the 1 percent level: ferent characteristics than verdicts, information on set- HOSPITAL, CHRONIC, and DEATH (model tlements is critical to understanding foodborne illness 2 2 R =0.5283; adj. R =0.4211). These findings make litigation. Future research may provide evidence that intuitive sense as once it has been decided that the the strongest incentives from the legal system for food plaintiff will prevail, the injury severity is a determi- firms to improve food safety are from the threat of nant of award size. These significant effects accord large outbreaks with associated widespread litigation well with previous research on award size (GAO 1989; and from the threat of uninsured economic losses. Rodgers 1993; Tabarrok and Helland, 1999) and suggest that juries try to fairly compensate plaintiffs for their losses. Meanwhile, claimed emotional distress, loss of consortium, and pain and suffering were not significantly influential in determining damages, perhaps because of the subjective nature of these claims. The nonsignificant effect of both DEEPPOCK and CHILD also fits this picture: juries are compensating for a specific injury, regardless of whether or not the plaintiff was a child or if the defendant had deep pockets.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 23 Chapter 5

Summary

Product liability is a seemingly powerful mechanism or know them. As a result, many consumer plaintiffs for shifting the costs of foodborne illness from the per- apparently entered court without strong evidence of a sons who become ill to the firms responsible for the causal link between their illness and the defendants contaminated product. However, high transaction and food product. information costs combined with the structure of the 4. The implicated pathogen may also be associated legal system limit the effectiveness of the litigation for with multiple foods or may be spread via other routes compensating ill consumers and providing firms with (e.g., person-to-person contact) (Draper, 1994), signals to produce safer food. This report has nine key increasing the number of potential causes of illness points, drawn both from the background literature re- and making it more difficult to prove that the cause view and from findings from the jury verdict analysis. was contaminated food from a particular firm. Point 1: Litigation for foodborne illnesses attributed 5. Many processed food products include a variety of to microbial contamination of food by firms is limited ingredients from different sources, further increasing because of high information and transaction costs. the difficulty of identifying the particular food item Reliable estimates of the annual number of foodborne (and firm) that caused the illness. illness claims and lawsuits are unavailable. However, 6. Most foodborne illnesses are not part of a well- a review of the evidence suggests that the vast majority identified foodborne disease outbreak, limiting the of foodborne illnesses attributed to microbial contami- availability of supporting information from public nation of food by firms do not result in foodborne ill- health authorities. ness lawsuits. We found that those stricken by a foodborne illness face high information and transaction 7. The complexity and slow pace of the legal system costs when pursuing legal compensation. In particular, impose high transaction and information costs for con- we uncovered several reasons that help explain why sumers who decide to pursue legal action. The aver- legal action is rare: age lawsuit in our data was not tried until more than 3 years after the incident that resulted in illness. 1. The incubation period between the consumption of contaminated food and the initial symptoms of food- 8. Some attorneys may be unwilling to take a foodborne illness makes it difficult for consumers to identi- poisoning case because: (a) they believe that a food fy the specific food item (and firm) which caused their product sample containing a pathogen is needed to illness. proceed with a case (untrue), (b) they are not educated about how to proceed with foodborne litigation, and 2. Physical evidence of contamination is rarely avail- (c) they rely only on court data and therefore think that able because most food items are likely to have been foodborne illness cases are not worth much consumed or thrown out before the initial symptoms of (Rosenbaum, 2000). illness appear. Ill individuals rarely save a sample of food to test for contamination. Other aspects of foodborne illness may also hinder or discourage ill individuals from pursuing legal compen- 3. Medical evidence from clinical specimens identify- sation. In particular, a high proportion of foodborne ing a specific illness-causing pathogen is rare.18 Less illnesses are relatively mild and result in only minor than half (48 percent) of the foodborne illness lawsuits financial losses for individual consumers, reducing the implicated a specific foodborne pathogen, toxin, or ill- incentives to pursue individual legal action. And, most ness, presumably because the plaintiff did not specify consumers have health insurance, and many workers

18 However, science is increasingly supplying more and bet- cated pathogen in their bloodstream. Although the presence ter tools to link specific illnesses to specific food sources, of an antibody does not show exactly when the illness or and these tools are being applied more frequently. exposure occurred, it is one piece of evidence that can be Foodborne-illness victims may carry antibodies to the impli- used to meet the causation requirement (Clark, 2000).

24 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA have sick leave benefits, limiting their direct financial perfect information about the true costs of producing losses due to foodborne illness. unsafe food or about the frequency with which firms are actually sued for food contamination. Point 2: Foodborne illness claims that are settled confidentially are likely to have different characteris- Point 3: Plaintiffs are unlikely to win awards in food- tics than those that reach court verdicts. This implies borne illness jury trials. that confidential settlements distort legal incentives to It appears that relatively few foodborne illnesses are produce safer food. compensated either through court awards or out-of- Although we do not have data indicating what percent court settlements. Of the WLN sample of 175 food- of foodborne illness claims are resolved through settle- borne illness lawsuits ultimately resolved in court dur- ments, data on the broader category of product liability ing 1988-97, 31.4 percent resulted in some compensa- cases involving bodily injury indicate that 95 percent tion paid by firms. This suggests that most plaintiffs of claims are settled out of court, and only 5 percent who go to trial do not have a strong case. Lack of ultimately reach a court verdict (Viscusi, 1991). These convincing evidence on causation is likely to be the data suggest that even when foodborne illness claims key factor in the low success rate of plaintiffs in court. are pursued, most are resolved through settlements Point 4: Plaintiffs were more likely to win jury trials between plaintiffs and firms (or their insurers) before if they could link their illness to a specific pathogen, trial. and more severe illnesses tended to result in higher A review of the legal literature also suggests that in awards. general, there are real differences between claims that Multivariate analyses indicate that the odds of a plain- result in settlements and claims that result in court ver- tiff victory increased if a foodborne pathogen or illness dicts (Vidmar, 1997; Eisenberg, 1991). In essence, was specified and decreased if defendants had deep tried cases are not representative of all claims (Priest pockets or used medical expert testimony. This high- and Klein, 1984). For foodborne illness claims, the lights the importance of a plaintiffs being able to link same distinction appears to hold. Foodborne illness the illness to a specific foodborne pathogen in order to claims that are bona fide (without deceit or fraud) tend prevail in court. Higher awards were given when the to settle while claims that go to trial are typically those illnesses involved hospitalization, death, or chronic where there is a serious causation question or where complications. the amount of damages is disputed (Clark, 2000). Point 5: The expected monetary compensation from When complaints and lawsuits are settled confidential- a foodborne illness lawsuit provides only limited ly, direct economic signals from the legal system about incentives to pursue litigation. the costs of producing pathogen-contaminated food are usually restricted to the responsible firm and its insurer Most foodborne illnesses are relatively mild and do not (which may then decide to raise the firms premium for result in high medical costs or lost productivity costs. product liability coverage).19 In particular, if there is And of those illnesses that are more severe and result selection bias determining which lawsuits end up in in higher costs, some portion of these costs is likely to confidential settlements and which lawsuits end up in be covered by other parties such as health insurance court decisions, the representativeness of awards publi- and employers. Therefore, most people with food- cized in the media and the associated feedback to other borne illness have weak monetary incentives to take firms to produce safe food will be distorted. Whisper legal action to recover damages. information among firms also occurs, but perhaps more so in more serious cases involving high monetary Only a third of jury trials involving injuries due to awards, once again implying that firms do not receive pathogen-contaminated food products resulted in monetary compensation for the injured consumer. However, some of these consumers received substantial sums (mean award was $133,280), significantly 19 Insurers actually share information about losses, and use elevating the expected award ($41,888) above the this information to estimate risks and set premiums. So the median award ($25,560) (all in 1998 dollars). It is economic signals are sent indirectly to the insurance market possible that plaintiff lawyers might be misled by this and subsequently also to other firms when premiums are seemingly high expected award and accept weak raised. However, increases in premiums encourage food firms to produce safer food only if the firm links the premi- claims to pursue in the legal system. Consumer plain- um increase to its food safety record. tiffs seldom receive all of the awards because a sub-

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 25 stantial portion (typically one-third or more) is usually meats produced and sold by Bil Mar and Sara Lee used to pay legal fees and court costs. Out-of-pocket (FSnet, Aug. 27, 1999). medical expenses, lost productivity, and other costs One issue needing study is how legal incentives from incurred because of the illness reduce the net award to outbreaks differ by size of firm. Most firms are small the plaintiff even further. operations, not giant national entities. Small firms are Point 6: Foodborne illness costs are shared by many probably less prepared to deal with legal consequences sectors of the economy, in turn limiting incentives to of foodborne illness than are large firms due to differ- firms to produce safer food. ences in insurance coverage, economic resources, size of market, in-house legal and disaster management Although we do not have reliable estimates of the total expertise, and so forth. If small firms have limited costs of illness to all sectors that are caused by safety insurance, they may have much higher incentives to lapses by food firms (as opposed to errors by con- see that claims are resolved because of the risk that sumers), the estimate is likely to be substantial. And, claims will outstrip coverage or that even modest local we suspect that food firms pay directly for only a publicity will hurt near-term business (Clark, 2000). small, unknown, portion of the total costs of illnesses caused by their errors. For example, much of the costs It is primarily the business disruption and negative of illness borne by people who became ill (and/or their publicity of the catastrophic foodborne illness or out- families) are not reimbursed by the food firms respon- breaks that cost firms money so it is these extraordi- sible. Instead, they are paid for by the ill consumer or nary, nonrecurrent illnesses or outbreaks that have the his or her household, shifted to other parties such as potential to substantively shape corporate behavior employers, private health insurers, and governments (Clark, 2000). In the rare instances where foodborne (and in turn, taxpayers), or handled by some combina- disease outbreaks are linked to particular firms, the tion of these parties. impact on those firms can be large. For example, Foodmaker Inc. [now Jack in the Box Inc.] lost an esti- Because a large share of the foodborne illness costs are mated $160 million in the first 18 months after the borne by consumers who become ill or by other sectors 1993 E. coli O157:H7 outbreak (Roberts et al., of the economy, firms receive only limited feedback to 1997).21 produce safer food. If food firms have sufficient product liability insurance to cover a lawsuit, they may not Point 8: It is unclear whether foodborne illness liti- feel the full financial impact, even if their premiums gation will become more common in the future. and those of similar firms should increase. One impli- However, class action lawsuits may become more cation of the current social allocation of foodborne ill- common in the case of outbreaks where many per- ness costs is that food firms probably underinvest in sons have similar, mild illnesses. food safety. It is unclear whether foodborne illness litigation will Point 7: Legal incentives probably work better in out- become more common in the future. Foodborne ill- break situations and less well for sporadic cases. nessand the reasons for litigationmay decrease if firms continue to improve quality control practices to Incentives for firms to avoid foodborne illness out- ensure safer food. In contrast, improvements in breaks are probably stronger than the incentives to pathogen detection and identification techniques avoid isolated, sporadic cases of illness because out- (including DNA fingerprinting and more rapid micro- breaks have greater potential to damage firms. Public bial tests) may increase the chances that foodborne ill- health authorities are also more likely to become nesses (particularly outbreaks) will be detected and involved in outbreaks and technological advances have linked to specific food products and firms. Attorneys improved the chances that widely scattered cases will who specialize in personal injury cases may also be traced back to a source and linked to each other.20 For example, CDC traced the 1998 listeriosis outbreak (80 illnesses, 21 deaths) to hot dogs and luncheon

21 Also, in the 6 months following the 1998 recall of Sara Lee Corp. hot dogs and deli meats due to contamination with Listeria, sales of Sara Lee meat products fell by about 20 This may mean that there will be an increase in the num- $200 million and the companys stock price fell by 19 per- ber of small outbreaks detected and a drop in the average cent, reducing the companys value from $25 billion to size of reported outbreaks. $20.3 billion (FSnet, Aug. 27, 1999).

26 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA become more interested in handling foodborne illness attention. Other firms may decide to increase invest- litigation as scientific and technological advances ments in food safety after observing the economic make it easier to link foodborne illnesses to individual costs to defendant firms accused of producing contam- firms. In addition, increasing consumer awareness that inated food products that caused foodborne illness. food products may be contaminated by pathogens Economic costs from these lawsuits that food firms could increase the likelihood of a claim after a food- wish to avoid include: (1) the potentially high legal borne illness occurs (Clark, 2000). The experience of costs and court fees involved in defending lawsuits, (2) Marler-Clark (the law firm that has handled more the compensation payments and possible punitive dam- foodborne illness litigation in the United States than ages when the defendant firm is found liable, and (3) any other law firm) is that consumers are increasingly business losses as a result of trial publicity. These interested in exploring legal recourse when made sick business losses include the loss of reputation of firm or by food (Clark, 2000). But, these trends may encour- product, reduced product demand, reduced stock age food firms to further improve quality control stan- prices, higher premiums for product liability insurance, dards to reduce the risk of producing contaminated temporary plant closings for cleanup, or permanent food products that might cause illness and result in liti- plant closings following adverse publicity about a gation. foodborne illness lawsuit, even when the firm success- Several law and consulting firms now specialize in fully defends itself. Catastrophic financial losses for foodborne illness lawsuits. Class action or mass defendants may result even when the law is apparently lawsuits may be more frequently used in the case of in their favor. In addition to the possibility of incur- outbreaks resulting in many similar, mild illnesses, ring these economic costs, the uncertain outcome in particularly as identification and documentation of out- the case of jury trials may be daunting, particularly for breaks improves, as legal expertise in this area grows, more risk-averse firms and firms with lower equity. and as media coverage of successful class action suits For example, an effective, industry-generated, food involving consumer products accumulates. safety reform occurred after the large 1993 outbreak Point 9: The legal system provides incentives, though from hamburgers contaminated with E. coli O157:H7 limited, for firms to produce safer food. and subsequent litigation. Jack in the Box, Inc., revamped its food safety program and significantly Because firms responsible for the microbial contami- altered the practices of the fast food industry with nation compensate relatively few foodborne illnesses, respect to protein products (Clark, 2000). As we are the legal system provides only limited feedback to increasingly able to identify the source of a foodborne firms about the need for greater food safety. The prod- illness, the power of litigation to shape industry behav- uct liability system provides firms with incentives to ior about food safety will increase (Clark, 2000). control hazards in food primarily when the hazards are easily identifiable, the foodborne illness can be traced Future research should focus on developing a better back to firms, and ill individuals or their families are understanding of the litigation process because food- compensated by the firms responsible for the contami- borne illness lawsuits are a potentially important eco- nation. These findings suggest that the direct impact nomic signal to firms to invest more in food safety. of litigation on firms is small, although few if any Specific questions for research include determining: firms are likely to ignore the potential legal conse- (1) how often lawsuits are filed, (2) how often lawsuits quences of making or distributing contaminated food are settled or otherwise resolved before trial, and (3) products that might cause illness or death. And, firms how settlements differ from court decisions. This cannot ignore the risk that they may face catastrophic information about noncourt cases is critical to improve losses if they produce contaminated food. It is diffi- understanding of the extent of foodborne illness law- cult to assess exactly how firms are affected by such suits and the true probability that plaintiffs will be legal action because the actual decisionmaking process compensated for damages caused by foodborne illness. on food safety issues by firms is generally kept confi- Consumer complaints and out-of-court settlements are dential. far more frequent than lawsuits that go to trial and may be the most common signals about the costs of unsafe However, the small percentage of foodborne illness food received by firms. jury trials that are resolved in the public view may have an indirect, possibly significant, impact on the behavior of the defendants and other firms. This is particularly true for lawsuits that attract adverse media

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 27 References

Adams, C. E. ISO 9000 and HACCP Systems. Food United States, 1988-1992. CDC Surveillance and Drug Law Journal. 49(1994):603-7. Summary. MMWR 45,SS-5(Oct. 25, 1996). Altekruse, S., M.L. Cohen, and D.L. Swerdlow. Benenson, A. S. (ed.), Control of Communicable Emerging Foodborne Diseases. Emerging Infectious Diseases in Man. Washington, DC: American Public Diseases. 3,3(July-Sept. 1997):285-93. Health Association, 15th edition, 1990. American Law Institute. Restatement of the Law, Berkelman, R. L., R. T. Bryan, M. T. Osterholm, J. W. Second, Torts 2d, American Law Institute Publishers, LeDuc, and J. M. Hughes. Infectious Disease St. Paul, MN, 1965. Surveillance: A Crumbling Foundation. Science. 264, 5157(April 15, 1994):368-70. American Law of Products Liability 3d, The Lawyers Co-Operative Publishing Co., Rochester, NY, 1987, § Black H. C. Blacks Law Dictionary. Sixth Edition. St. 81.1 and § 81.3. Paul, MN: West Publishing Co., 1990. American Public Health Association et al. v. Earl Butz, Bonner, P. A., and E. J. Metzen. Small Claims Court: Secretary of Department of Agriculture, et al., 511 Factors Influencing a Successful Outcome for the Federal Reporter, 2d series, 331-8, 1974. Individual Plaintiff. The Journal of Consumer Affairs. 26,1(Summer 1992):1-19. Applebaum, R. Introduction in V. Quesada, ed., Strategies for Defending Food Litigation, National Bradgate, J. R., and G. G. Howells. Food Law in the Food Processors Association. Washington, DC, Oct. United Kingdom. Food and Drug Cosmetic Law 1995. Journal. 17(May 1991):447-66. Archer, D. L., and J. E. Kvenberg. Incidence and Cost Broder, I. E. Characteristics of Million Dollar of Foodborne Diarrheal Disease in the United States. Awards: Jury Verdicts and Final Disbursements Journal of Food Protection. 48(1985):887-894. Justice System Journal. 11(1986):349-59. Atiyah P. S. Tort Law and the Alternatives: Some Broder, I., and J. F. Morrall III. Incentives for Firms Anglo-American Comparisons. Duke Law Journal. to Provide Safety: Regulatory Authority and Capital 6(1987):1002-44. Market Reactions. Journal of Regulatory Economics. 3(1991):309-22. Bailis, D. S., and R. J. MacCoun. Estimating Liability Risks with the Media as Your Guide: A Content Bryan, F. L., J. J. Guzewich, and E. C. Todd. Analysis of Media Coverage of Tort Litigation. Law Surveillance of Foodborne Disease III. Summary and and Human Behavior. 20,4(Aug. 1996):419-429. Presentation of Data on Vehicles and Contributory Factors; Their Value and Limitations. Journal of Food Barker, J. P. Taking A Byte Out of Abusive Agency Protection 60,6(1997):701-14. Discretion: A Proposal for Disclosure In the Use of Computer Models. University of Michigan Journal of Burditt, G., M. The History of Food Law. Food and Law Reform. 19,3(Spring 1986):637-668. Drug Law Journal. 50(1995):197-201. Bean, N. H., and P. M. Griffin. Foodborne Disease Buzby, J. C. and S. Crutchfield. New Juice Outbreaks in the United States, 1973-1987: Pathogens, Regulations Underway. U.S. Dept. of Agr., Econ. Res. Vehicles, and Trends. Journal of Food Protection. Serv., FoodReview. 22,2(May-Aug. 1999):23-25. 53(1990):804-817. Buzby, J. C., J. A. Fox, R.C. Ready, and S. R. Bean, N. H., P. M. Griffin, J. S. Goulding, and C. B. Crutchfield. Measuring Consumer Benefits of Food Ivey. Foodborne Disease Outbreaks, 5 Year Summary, Safety Risk Reductions. Journal of Agriculture and 1983-1987. CDC Surveillance Summary. MMWR Applied Economics. 30,1(July 1998):69-82. 39,SS-1(Mar. 1990):15-59 and Journal of Food Buzby, J. C., and P. Frenzen. Product Liability and Protection. 53(1990):711-728. Food Safety, Food Policy. 24(1999): 637-651. Bean, N. H., J. S. Goulding, C. Lao, and F. J. Angulo. Surveillance for Foodborne Disease Outbreaks -

28 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA Buzby, J. C., and T. Roberts. Economic Costs and Danzon, P. M. Liability for Medical Malpractice. Trade Impacts of Foodborne Illness. World Health Journal of Economic Perspectives. 5,3(Summer Statistics Quarterly. 50, ½(1997):57-66. 1991):51-69. Buzby, J. C., T. Roberts, C. -T. J. Lin, and J. M. Daughety, A. F., and J. F. Reinganum. Product Safety: MacDonald. Bacterial Foodborne Disease: Medical Liability, R&D, and Signaling. The American Costs and Productivity Losses. U.S. Dept. of Agr., Economic Review. 85,5(Dec. 1995):1187-1206. Econ. Res. Serv., AER-741, Aug. 1996. Draper, J. M. Liability for Injury of Death Allegedly Calfee J. E., and C. Winston. The Consumer Welfare Caused by Spoilage, Contamination, or Other Effects of Liability for Pain and Suffering: An Deleterious Condition of Food or Food Product. Exploratory Analysis. Brookings Papers on Lawyers Coop. Pub., 1994 (2 ALR 5th 1). Economic Activity (BPEA): Microeconomics. No. 1, Eads, G., and P. Reuter. Designing Safer Products: 1993, Washington, DC: The Brookings Institution. Corporate Responses to Product Liability Law and Council for Agricultural Science and Technology Regulation. Journal of Products Liability. (CAST) Report, Foodborne Pathogens: Risks and 7(1984):263-94. Consequences. R. 122, Ames, IA, Sept. 1994. Eisenberg, T. The Relationship between Plaintiff Caswell, J.A., and N. H. Hooker. HACCP as an Success Rates Before Trial and at Trial. Journal of the International Trade Standard. American Journal of Royal Statistical Society. 154,Part 1(1991):111-6. Agricultural Economics. 78,3(Aug. 1996):775-9. Eisenberg, T., and K. M. Clermont. Internet database Caswell, J. A., and G. V. Johnson. Firm Strategic on federal district-court civil cases terminated since Response to Food Safety and Nutrition Regulation. 1978. . [Feb. 15, 2000]. Economics of Food Safety, (ed.) J. Caswell. New York: Farnsworth, E. A. An Introduction to the Legal System Elsevier Science Pub. Co., (1991):273-297. of the United States. 2nd Ed. London: Oceana Clark, Bruce T. Attorney at Marler Clark. Personal Publications Inc., 1987. Correspondence, 2000. Fein, S. B., C.-T. J. Lin, and A. S. Levy. Foodborne Clermont, K. M., and T. Eisenberg. Trial by Jury or Illness: Perceptions, Experience, and Preventive Judge: Transcending Empiricism. Cornell Law Behaviors in the United States. Journal of Food Review. 77(July 1992):1124-1177. Protection. 58,12(Dec. 1995):1405-11. Cooper, R. M. Regulation through Private Litigation. Fein, S. B., and T. A. Riggins. Trends in Food Safety Food and Drug Law Journal. 47(1992):437-44. Practices, Concern, and Awareness 1988, 1993, and 1998. Presented at the Food and Drug Admin.-U.S. Cooter, R. D. Economic Theories of Legal Liability. Dept. of Agric. Food Safety Educ. Working Group Journal of Economic Perspectives. 5,3(Summer meeting, July 28-29, 1998, Washington, DC. 1991):11-30. Fidler, D. P. Legal Issues Associated with Cox, D. R., and E. J. Snell. The Analysis of Binary Antimicrobial Drug Resistance. Emerging Infectious Data. Second Edition, London: Chapman and Hall, Diseases. 4,2(April-June, 1998):169-177. 1989. Food Chemical News, March 29, 1982, Nov. 21, 1994, Council of Economic Advisors. Executive Office of Sept. 4, 1996, Sept. 16, 1996, and Mar. 24, 1997, the President. Economic Report of the President: Washington DC: CRC Press LRC. Transmitted to the Congress, Feb. 1996, Together with the Annual Report of the Council of Economic Franklin, T. J. Calculating Damages for Loss of Advisors, Feb. 1996, ISBN 0-16-048501-0. Parental Nurture through Multiple Regression Analysis. Washington and Lee Law Review. Crutchfield, S., and T. Roberts. Food Safety Efforts 52(1995):271-320. Accelerate in the 1990s. U.S. Dept. of Agr., Econ. Res. Serv. FoodReview. 23,3(Sept.-Dec. 2000): 44-49. Frenzen, P., T. Riggs, J. Buzby, T. Breuer, T. Roberts, D. Voetsch, S. Reddy, and the FoodNet Working Daniels, S., and J. Martin. Jury Verdicts and the Group. Salmonella Cost Estimate Update Using Crisis in Civil Justice. Justice System Journal. 11(1986):321-348.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 29 FoodNet Data. U.S. Dept. of Agr., Econ. Res. Serv., United States. 1991. R-3999-HHS/ICJ. RAND, Santa FoodReview, 22,2(May-Aug. 1999):10-15. Monica, CA. Frumer, L. R., and M. I. Friedman. Products Liability. Hutt, P. B. Food Technology and Food Law: Uneasy New York: Matthew Bender, 1986. Partnership. Food Technology. 39(1985):14-18. Food Safety Network (FSnet). May 14, 1996, Sept. 26, Institute of Medicine. Seafood Safety. Committee on 1996, July 3, 1997, July 7, 1997, and Aug. 27, 1999 Evaluation of the Safety of Fishery Products. issues. University of Guelph, Canada, Landover, MD, 1991. . [Jan. 15., Institute of Food Technologists Expert Panel on Food 2001]. Safety and Nutrition. Scientific Status Summary, Garber, S. Product Liability, Punitive Damages, Foodborne Illness: Role of Food Handling Practices. Business Decisions, and Economic Outcomes. Food Technology. 49(1995):119-31. Wisconsin Law Review 1998,1(1998a):237-295. Jackson, G. J. Principles and Costs in the Regulation Reprinted by Rand, Institute for Civil Justice, of Microbially Contaminated Foods. Regulation of RAND/RP-747. Contaminated Food. (1990):382-3. Garber, S. Good Deterrence, Bad Deterrence and Johnson, G. V., T. S. Ulen, and J. M. Alho. Optimal Challenges in Product Liability Reform. Rand Public Policies to Minimize Risks of Genotoxic Harms: Institute for Civil Justice (1998b). a Comparison of the Efficiency of Administrative . [July Agency Regulation & Tort Liability. Food Marketing 31, 1998]. Policy Center, Res. Rep. No. 1, Dept. of Agr. Econ. & Gates v. Standard Brands Inc., 43 Wash. App. 520, Rural Soc., Storrs, CT: University of Connecticut, 1986, 719 P.2d 130. 1989. Gifis, S. H. Law Dictionary. Woodbury, NY: Barrons Keeton, W. P., and D. B. Dobbs, R. E. Keeton, and D. Education Series, Inc., 1975. G. Owen. (eds.) Prosser and Keaton on The Law of Torts. 5th edition. St. Paul, MN: West Publishing Co., Golan, E. H., K. L. Ralston, and P. D. Frenzen. A 1984. Distributional Analysis of the Costs of Foodborne Illness: Who Ultimately Pays? Journal of Agricultural Kilpatrick, William v. The Superior Court of the City and Applied Economics. 30,1(July 1998):95-107. and County of San Francisco, Respondent. Holiday Inns, Inc., et al., Real Parties in Interest. 11 Cal. Gold, S. Food Law, Glorious Food Law. The New Rptr.2d 323 (Cal. App. 1 Dist. 1992). Law Journal. 141(Feb. 1991):158. Kolstad, C. D., T. S. Ulen, and G. V. Johnson. Ex Post The Guide to American Law: Everyones Legal Liability for Harm vs. Ex Ante Safety Regulation: Encyclopedia, St. Paul, MN: West Publishing Co., Substitutes or Complements? The American 1984, p. 320. Economic Review. 80,4(Sept. 1990):888-901. Harl, N. E. Agriculture Law. Through release # 50. Kumamoto, H., and E. J. Henley. Probabilitistic Risk New York: Matthew Bender, 1997. Assessment and Management for Engineers and Helmick, C. G., P. M. Griffin, D. G. Addiss, R. V. Scientists. New York: The Institute of Electrical and Tauxe, and D. D. Juranek. Chapter 3 in T. W. Electronics Engineers, Inc., 1996. Hennessy, C. W. Hedberg, L. Slutsker, et al. National Landa, M. M. Regulatory Programs for Naturally Outbreak of Salmonella enteritidis Infections from Ice Occurring Toxicants. Food and Drug Law Journal. Cream. New England Journal of Medicine. 334, 48(Jan. 1993):125-35. 20(May 16, 1996):1281-6. Landes, W. M., and R. A. Posner. The Economic Hensler, D. R. Reading the Tort Litigation Tea Structure of Tort Law. Cambridge, MA: Harvard Leaves: Whats Going on in the Civil Liability University Press, 1987. System? The Justice System Journal. 16,2(1993):139- 154. Lederman, L. Which Cases Go to Trial? An Empirical Study of Predictors of Failure to Settle. Case Western Hensler, D. R., M. S. Marquis, A. F. Abrahamse and Reserve Law Review. 49(Winter 1999):315-358. others. Compensation for Accidental Injuries in the

30 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA Levit, K. R., and M. S. Freeland. National Medical McNamara, S. H. A Legal Assessment of FDAs Care Spending. Health Affairs. (Winter 1988):124- HACCP Regulations. Food and Drug Law Journal. 136. 52,1(1997):39-47. Lindsay, J. A. Chronic Sequelae of Foodborne Mennell, P. S. The Limitations of Legal Institutions Disease. Emerging Infectious Diseases. 3,4(Oct.-Dec. for Addressing Environmental Risks. Journal of 1997):443-52. Economic Perspectives. 5,3(Summer 1991):93-113. Litan, R. E. The Safety and Innovation Effects of Mergenhagen, P. Product Liability: Who Sues? U.S. Liability Law. The American Economic Review. American Demographics. 17(1995):48-55. 81,2(May 1991):59-64. . Long, J. S. Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks, CA: Merritt, D. J., and K. A. Barry. Is the Tort System in Sage Publications, Inc., 1997. Crisis? New Empirical Evidence. Ohio State Law Journal. 60,2(1999):316-398. Low, S., and J. K. Smith. Decisions to Retain Attorneys and File Lawsuits: An Examination of the Moller, E. 1996. Trends in Civil Jury Verdicts Since Comparative Negligence Rule in Accident Law. 1985. Report No. MR-694-ICJ. Santa Monica, CA: Journal of Legal Studies. 24(June 1995):535-557. RAND, Inc., 1996. Madden, J. M. Concerns Regarding the Occurrence of Mookherjee, D., and I. P. L. Png. Monitoring Vis-à- Listeria monocytogenes, Campylobacter jejuni, and vis Investigation in Enforcement of Law. The Escherichia coli O157:H7 in Foods Regulated by the American Economic Review. 82,3(June 1992):556-565. U.S. Food and Drug Administration. Dairy, Food, and Munarriz, R. A. Odwalla, Inc. Daily Trouble, Nov. Environmental Sanitation. 14,5(May 1994):262-7. 25, 1997. . [Oct. 14, 1998]. International Standards for Food Production and Murphy, S. L. Deaths: Final Data for 1998. National Safety. Journal of Agricultural Taxation and Law. Vital Statistics Reports; Vol. 48, No. 11. Hyattsville, 12(Winter 1991):334-41. MD: National Center for Health Statistics, 2000. Mann, J.M., G.D. Lathrop, and J.A. Bannerman. Nagelkerke, N. J. D. Miscellanea: A Note on a Economic Impact of a Botulism Outbreak: General Definition of the Coefficient of Importance of the Legal Component in Food-Borne Determination. Biometrika. 78,3(1991):691-2. Disease. JAMA. 249,10(Mar. 11, 1983):1299-1301. Olsen, S. J., L. C. MacKinon, J. R. Goulding, N. H. Manoukian, N. The Federal Governments Inspection Bean, and L. Slutsker. Surveillance of Foodborne and Labeling of Meat and Poultry Products: Is it Disease Outbreaks, United States--1993-97. MMWR, Sufficient to Protect the Publics Health, Safety and Surveillance Summary. 49,SS01(Mar. 17, 2000):1-51. Welfare? Western State University Law Review. 21(1994):563-77. Parry, G. J. Guillain-Barré Syndrome. New York: Thieme Medical Publishers, Inc., 1993. Mayer, C. E. Another Round of Tort and Re-Tort: This Time, Measure Curbing Product Liability Suits Polinsky, A. M. Strict Liability vs. Negligence in a Has White House Backing. The Washington Post, Market Setting. Borderlines of Law and Economic Tues. July 7, 1998, pp. E1-E2. Theory. 70,2(May 1980):363-367. Mead, P. S., L. Finelli, M. A. Lambert-Fair, et al. Risk Posner, R. A. Explaining the Variance in the Number Factors for Sporadic Infection with Escherichia coli of Tort Suits Across U.S. States and Between the O157:H7. Archives of Internal Medicine. United States and England. Journal of Legal Studies. 157(1997):204-8. 26(June 1997):477-488. Mead, P. S., L. Slutsker, V. Dietz, L. F. McCaig, J. S. Posner, R. A. Law and Legal Theory in the UK and the Bresee, C. Shapiro, P. M. Griffin, and Robert V. Tauxe. USA, Oxford: Clarendon Press, 1996. Food-Related Illness and Death in the United States. Emerging Infectious Diseases. 5,5(Sept.-Oct. Post, R. C. Keeping Food Safe. Food and Drug Law 1999):607-625. Journal. 50(1995):389-93.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 31 Powell, D. Food Safety Claims Done in by Data. Rodgers, G. B. Estimating Jury Compensation for Farm and Country. 63:16 (May 3 1999):4. Pain and Suffering in Product Liability Cases Involving Nonfatal Personal Injury. Journal of Poyck, E. A., M. A. Smith, and A. Toledo. Federal Forensic Economics. 6,3(1993):251-261. Food and Drug Act Violations. American Criminal Law Review. 30(1993):717-31. Rose-Ackerman, S. Regulation and the Law of Torts. The American Economic Review. 81,2(May 1991):54- Priest, G. L. The Modern Expansion of Tort Liability: 58. Its Sources, Its Effects, and Its Reform. Journal of Economic Perspectives. 5,3(Summer 1991):31-50. Rosenbaum, D. at Food Safety Partners, Northbrook IL, personal conversation, July 1998 and Aug. 2000. Priest, G. L., and B. Klein. The Selection of Disputes for Litigation. Journal of Legal Studies. 13(Jan. Rubin, R. I., and K. E. Lamb. Tainted Food: A Brief 1984)1-55. Digest of Liability Theories. Trial. 29(Mar. 1993):128-34. Quesada, V. Products Liability: Basic Theory and Practice, Chapter 1 in V. Quesada, ed., Strategies for Schultz, H. W. Food Law Handbook. Westport, CN: Defending Food Litigation, National Food Processors The AVI Publishing Co. Inc., 1981. Association. Washington, DC, Oct. 1996. The San Diego Union-Tribune, San Diegos RAND Institute for Civil Justice Frequently Asked Foodmaker Faces Class-Action Stockholders Suit, Questions. http://www.rand.org/ Associated Press. pp. C-3, April 17, 1993. center/icj/faqs/ans1/.html) accessed July 31, 1998 Scott, E., and P. Sockett. How to Prevent Food (1998a). Poisoning: A Practical Guide to Safe Cooking, Eating, RAND Institute for Civil Justice Escaping the and Food Handling. New York: John Wiley & Sons, Courthouse. (1998b)(http://www.rand.org/publica- 1998. tions/RB/RB90205/RB9020.doc.html). Segerson, K. Liability and Penalty Structures in RAND Institute for Civil Justice Understanding Mass Policy Design. Chapter 13 in D.W. Bromley, ed., The Personal Injury Litigation. Handbook of Environmental Economics. Malden, MA: (1998c)(http://www.rand.org/publications/RB/RB9021/ Basil Blackwell, 1995. RB9021.word.html). Shanley, M. G. The Distribution of Posttrial RAND Institute for Civil Justice Trends in Civil Jury Adjustments to Jury Awards. Journal of Legal Verdicts: New Data from 15 Jurisdictions. Studies. 20(June 1991):463-481. (1998d)(http://www.rand.org/publications/RB/RB9025/ Shapiro, C. Symposium on the Economics of RB9025.html). Liability. Journal of Economic Perspectives. Rasco, B. How HACCP Can Help in a Products 5,3(Summer 1991):3-10. Liability Action. Food Quality. (Oct. 1997):16-8. Shavell, S. Economic Analysis of Accident Law. Riggs, R. Jack In The Box Franchise Holders Sue Cambridge, MA: Harvard Univ. Press, 1987. San Diego Union-Tribune, Friday, July 9, 1993. Silverman, R. S. Legal Implications of Risk Roach, J. Odwalla Makes a Comeback. Assessments in Food Processing Operations. Food Environmental News Network, Sept. 21, 1999. Technology. (Dec. 1996):65-67. , accessed Claims, Chapter 2 in V. Quesada, ed., Strategies for Nov. 1, 1999. Defending Food Litigation, Washington, DC: National Roberts, T., R. A. Morales, C.-T. J. Lin, J. A. Caswell., Food Processors Association, 1996. and N. H. Hooker. Worldwide Opportunities to Slutsker, L., A. A. Ries, K. Maloney, J. G. Wells, K. D. Market Food Safety. Ch. 10 in Government and the Greene, and P. M. Griffin. A Nationwide Case- Food Industry: Economic and Political Effects of Control Study of Escherichia coli O157:H7 Infection Conflict and Cooperation, eds. L. T. Wallace and W. R. in the United States. Journal of Infectious Diseases. Schroder, Boston: Kluwer Academic Publishers, 1997. 177,4(April 1998):962-6.

32 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA Steenson, M. A Comparative Analysis of Minnesota U.S. Centers for Disease Control and Prevention Products Liability Law and the Restatement (Third) of (CDC), Surveillance for Foodborne-Disease Torts: Product Liability. William Mitchell Law Outbreaks--United States, 1988-1992. MMWR 45,SS- Review. 24,1(1998). 5(Oct. 25, 1996):58-66. Tabarrok, A., and E. Helland. Court Politics: The U.S. Consumer Product Safety Commission. 1997 Political Economy of Tort Awards. Journal of Law Annual Report to Congress. 1998. and Economics. 41,1(April 1999):157-188. U.S. Food and Drug Administration, Center for Food Tarr, P. I., N. T. Tran, R. A. Wilson. Escherichia coli Safety and Applied Nutrition. Foodborne Pathogenic O157:H7 in Retail Ground Beef in Seattle: Results of a Microorganisms and Natural Toxins Handbook, Bad One-Year Prospective Study. Journal of Food Bug Book http://vm.cfsan.fda.gov/~mow/appl.html. Protection. 62,2(Feb. 1999):133-9. [Dec. 1991]. Tauxe, R. V. Emerging Foodborne Diseases: An U.S. General Accounting Office. Federal Employees Evolving Public Health Challenge. Emerging Liability Act: Issues Associated with Changing How Infectious Diseases. 3,4(Oct.-Dec. 1997):425-434. Railroad Work-Related Injuries are Compensated. GAO/RCED-96-199, Aug. 15, 1996. Taylor, M. R. Preparing Americas Food Safety System for the Twenty-First Century--Who is U.S. General Accounting Office. Product Liability: Responsible for What When it Comes to Meeting the Verdicts and Case Resolution in Five States. GAO Food Safety Challenges of the Consumer-Driven HRD-89-99. Sept. 1989. Global Economy? Food and Drug Law Journal. Vidmar, N. Medical Malpractice and the American 52(1997):13-30. Jury: Confronting the Myths about Jury Incompetence, Tobolowsky, P. M., and J. F. Quinn. Drug-Related Deep Pockets, and Outrageous Damage Awards. Ann Behavior as a Predictor of Defendant Pretrial Arbor: University of Michigan Press, 1997. Misconduct. Texas Tech Law Review. 25(1994):1019- Viscusi, W. Kip. Economic Foundations of the 1053. Current Regulatory Reform Efforts. Journal of Todd, E. C. D. Legal Liability and its Economic Economic Perspectives. 10,3(Summer 1996):119-134. Impact on the Food Industry. Journal of Food Viscusi, W. Kip. The Value of Risks to Life and Protection. 50(Dec. 1987):1048-57. Health. Journal of Economic Literature. 31(Dec. Todd, E. C. D. Economic Loss from Foodborne 1993):1912-46. Disease and Non-Illness Related Recalls Because of Viscusi, W. Kip. Product and Occupational Liability. Mishandling by Food Processors. Journal of Food Journal of Economic Perspectives. 5,3(Summer Protection. 48,7(July 1985a):621-33. 1991):71-91. Todd, E. C. D. Economic Loss from Foodborne Viscusi, W. K. Toward a Diminished Role for Tort Disease Outbreaks Associated with Foodservice Liability: Social Insurance, Government Regulation, Establishments. Journal of Food Protection. and Contemporary Risks to Health and Society. Yale 48,2(Feb. 1985b):169-180. Journal on Regulation. 6,1(Winter 1989):5-107. Turner, A. Product Liability and the Manufacture of Viscusi, W. K. Pain and Suffering in Product Tort Wholesome Food. Proceedings of Food Science and Liability Cases: Systematic Compensation or Technology meeting. 16,1(Mar. 1983):184-97. Capricious Awards? International Review of Law and U.S. Bureau of the Census. Statistical Abstract of the Economics. 8(1988):203-220. United States: 1999, 119th edition. 1999. Weddig, L. J. Economics of HACCP for the Seafood U.S. Centers for Disease Control and Prevention Industry. Food and Drug Law Journal. 49(May (CDC). Epidemiologic Notes and Reports Update: 1994):493-8. Salmonella enteritidis Infections in the Northeastern United States. MMWR Weekly. 36,13(April 10, 1987):204-5.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 33 Appendix

Product Liability Law As It Applies to Foodborne Illness

The Legal Process and pays any resultant settlement or judgment (Clark, 2000).2 (For simplicity here, we use the term firms A person affected by foodborne illness who believes he to include both defendants and their insurers.) or she can identify the injurer (i.e., the party responsible for the foodborne illness) can attempt to obtain Comprehensive data on the proportion of foodborne compensation for his or her costs of the illness by illness claims that become lawsuits or the proportion directly contacting the injurer, contacting the injurers of foodborne illness lawsuits that reach different stages insurer, consulting an attorney about pursuing litiga- of the litigation process are unavailable. However, tion, or some combination of these actions (Hensler et what is clear is that most of the 76 million foodborne al., 1991). A claim is any effort, including the filing of illnesses in the United States each year never result in a lawsuit, by an individual or group to obtain compen- a foodborne illness claim. Among the claims that are sation, directly or through an attorney, for injuries or pursued, the vast majority is resolved without a trial illnesses suffered. This appendix provides some back- through settlements and other types of negotiations. ground on U.S. product liability law as it relates to Among the claims that become lawsuits, relatively few foodborne illness, a type of personal injury. reach the courtroom to be resolved in jury trials. Only a very small portion of jury verdicts is later overturned Most people pursuing foodborne illness litigation through post-trial motions or following appeals from (plaintiffs) hire attorneys as the first step, but the for- dissatisfied plaintiffs or defendants. mal litigation process begins once a document called a complaint is filed in a court, describing the illness, cit- Some legal experts believe that unlike some other ing laws that may have been violated, and identifying areas of law such as auto accident claims where there one or more parties or defendants allegedly responsible is a higher proportion of fraudulent claims, most food- for the illness. borne illness claims are bona fide (Clark, 2000), which means without deceit or fraud (Blacks Law l Plaintiffs are generally individuals seeking compen- Dictionary, 1975). And because most foodborne ill- sation for their foodborne illness but may also include ness claims are bona fide, they are resolved without a parents on behalf of dependent children, guardians on trial (Clark, 2000). Settlements occur when firms pay behalf of legally incompetent individuals, and estates compensation to plaintiffs prior to trial.3 Legal ana- seeking monetary damages in cases of wrongful death. lysts believe that clear cases of product liability settle Plaintiffs may also include food companies who sue relatively quickly while more complex cases go to trial suppliers or others in the food distribution chain for loss of reputation and goodwill, loss of profits, and other damages. 2 Litigation may be very costly for firms, even when they l Defendants are usually firms that produce, process, have insurance coverage (e.g., business disruption, negative distribute, or sell food products although they may publicity, reduced product demand). occasionally include individual proprietors, employees 3 Most cases are settled out of court through informal nego- such as food servers, or even hosts of informal meals tiations between the lawyers for each side. Other cases are or other events where food was served.1 Almost all settled through alternate dispute resolution (ADR) procedures such as mediation and arbitration. Mediation involves firms in the food industry have some form of insurance both parties agreeing to the selection of an independent third protection against a foodborne illness claim (Clark, party (mediator) who will help the parties resolve the issues 2000). When one of these firms is sued, the insurance between them, reach a settlement, and draft a mediation company provides a legal defense at the insurers cost agreement. This agrement is binding upon the parties and is enforceable by the courts. Arbitration ocurs when the parties, either by prior contact, or by court order, submit their dispute to an impartial person or group of people for resolu- 1 Employees will rarely be named individually in a lawsuit, tion. In this report, the term settlements is used to cover and then only for tactical reasons (Clark, 2000). They will all types of formal and informal negotiations through which face no individual liability (Clark, 2000). a lawsuit is resolved.

34 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA (Eisenberg, 1991). Foodborne illness claims that tend to publicize the largest settlements. Either way, become lawsuits are typically those where there is a the direct feedback and information to other firms serious causation question or where the amount of about the costs of producing safer foods is distorted damages is disputed (Clark, 2000). In the case of med- and limited. However, when insurers pay for a large ical malpractice lawsuits, Vidmar (1997) found that foodborne illness claim, they may raise the premium there is a selection process whereby the cases with rates of similar firms, in effect providing indirect feed- ambiguous evidence tend to be decided by juries. This back to firms about the costs of not producing safer pattern suggests that court cases are not representative food. Insurers may also drop firms that cause repeated of all claims (Priest and Klein, 1984). insurance losses. And big settlements always generate whisper information (i.e., private discussions) among Firms have incentives to settle and keep the terms con- firms, and repeat offenders face pressure from insurers fidential in order to limit legal costs, avoid the uncer- or risk losing market share to safer competitors (Clark, tain outcome of a public trial and delays to resolution, 2000). avert the potentially adverse impact of a public trial on the reputation of the firm or its products, and avoid encouraging copycat claims and lawsuits. National or U.S. Product Liability Law well-known food firms may be particularly likely to Is State Law settle cases quickly (Rosenbaum, 1998). Because defendants insurers are risk averse, settlements will Product liability law governs most legal actions arising always predominate in bona fide claims (Clark, 2000). from foodborne illness. This branch of tort law describes the circumstances under which one can An extensive 1989 survey of accidental personal recover damages for a defective food item. The laws of injuries by RAND found that liability claims were individual States govern the nature and extent of com- filed in 3 percent of accidents such as product-related pensation that may be awarded for injuries or deaths injuries, slips, and falls, excluding work-related and due to contaminated food products (Clark, 2000). motor vehicle accidents (Hensler et al., 1991). The There is no uniform and comprehensive Federal law claim rate may be even lower for foodborne illness governing product liability in the United States. State than for other types of personal injuries because the trial courts (e.g., District or Superior court) have juris- cause of injury is not immediately apparent and identi- diction over foodborne illness cases. However, some fication of the source of contamination is often diffi- foodborne illness cases are litigated in Federal courts cult or impossible. Although these data are for acci- (Clark, 2000). These cases are typically removed from dents, foodborne illness and accidents are similar in State court to Federal court when the defendant firm that the greater the injury, the greater the incentive to invokes diversity jurisdiction (Clark, 2000), such as sue. when the parties are from different States, or when one Earlier data from the 1977 Insurance Services Office party requests a trial in Federal court as a means to Product Liability Closed Claims Survey of 23 insur- minimize perceived biases which would negatively ance companies (n=10,784 claims) indicate that 19 impact the party if the case were tried locally. Most percent of claims against insurers were dropped by the plaintiff lawyers, however, prefer to represent their plaintiffs. An estimated 95 percent of other claims clients in State courts (Clark, 2000). were settled out of court, and only 5 percent ultimately The annual number of product liability cases is reached a court verdict by a jury or a judge (Viscusi unknown. Michael Saks, a University of Iowa law 1991). professor, estimates that there may be up to 90,000 The significant number of confidential settlements that product liability cases (all kinds, not just foodborne ill- keep the monetary value and other information on a ness) filed in State courts each year (Mergenhagen, case out of publicly available court records, legal sum- 1995). In 1993, there were 18,959 product liability maries, and the media suggests that the available data cases filed in Federal courts, up from 2,393 cases in on foodborne illness settlements underestimates the 1975 and 8,026 cases in 1983 (Mergenhagen, 1995). relative number of plaintiffs compensated through settlements. The average compensation may also be Causes of Action underestimated if cases with the potentially highest damages are more likely to be settled out of court and The complaint a plaintiff files in court must identify or kept confidential. On the other hand, the average com- specifically state one or more legally recognized caus- pensation may be overestimated if plaintiffs lawyers es of action, facts supporting the elements of a cause

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 35 of action, and a demand for damages or some other the product.5 One example is if a person with liver form of judicial relief. Buzby and Frenzen (1999) out- disease knows the risks of eating raw oysters yet still line the three main causes of action raised in foodborne continues to eat them and then becomes ill. Another illness lawsuits: strict product liability, negligence, and consideration in determining liability is whether or not breach of express or implied warranty (table 11, fig. 2). consumers can and did take precautionary measures Additionally, plaintiffs may also raise other issues such while handling and cooking food. The tapeworm, as misrepresentation (Quesada, 1995); however, these Trichinae, in raw pork does not make the pork defec- would not be separate claims. tive and unreasonably dangerous to consumers because consumers are generally aware of the risks of eating undercooked pork and know that they can elimi- Strict Product Liability nate this risk by cooking pork thoroughly.6 To recover monetary compensation from the defendant Liability for microbial contamination of restaurant under a strict product liability cause of action, the food does not fall on the consumer because the con- plaintiff must prove that the product was defective and sumer had little if any control over how the food was unreasonably dangerous when it left the manufactur- prepared. As an aside, liability for other hazards asso- ers control and that this defect proximately caused the ciated with restaurant foods is less clear cut. For injury (see Harl 1997, p. 7-17). Proximate cause is the example, a consumer in a restaurant situation is still legal term for the link connecting the illness or injury expected to examine the food for obvious physical haz- with the product defect. The defendants liability is ards, such as bones in a whole intact chicken leg. limited to only foreseeable damages suffered by However, the situation is less clear when a physical foreseeable plaintiffs.4 hazard, such as bone fragments are present in shredded In other words, the central issue in most strict product chicken, hamburger patties, or other foods.7 In these liability foodborne illness cases is causation, i.e., it is situations, liability will fall with the party who was in the plaintiffs burden to prove by a preponderance of the best position to examine the food and remove the the evidence (more than a 50 percent likelihood) that hazard. his illness resulted from the food item at issue (Clark, 2000). If strict liability is proven, the defendant is Negligence liable for damages. For example, the defendants care in manufacturing or handling the product is irrelevant Negligence in foodborne illness cases occurs when the under strict liability and is not considered a defense defendant fails to exercise reasonable care in produc- (Clark, 2000). The focus is not on whether the defen- ing, marketing, or selling the implicated food, and dant food firm did anything wrong, but simply whether because of this failure, someone became ill. The three the illness in question can be traced to its product. elements that the plaintiff must prove in order to Strict liability is usually unsuccessful in foodborne ill- recover compensation under this cause of action are: ness litigation. (1) the defendant had a legal duty to exercise reasonable care in manufacturing the product and to warn all One consideration by the courts is the recognition that most foods cannot be made risk-free. For example, Campylobacter is a naturally occurring bacteria found 5 in poultry that may contaminate poultry products American Law of Products Liability, 1987, p. 9, § 81:3. 6 despite intensive efforts to prevent, reduce, or elimi- Restatement of Torts (2d § 402A); American Law of Products Liability, 1987, p. 9, § 81:3. Also, see Trabaudo v. nate contamination. Therefore, the defendant may not Kenton Ruritan Club, Inc. (517 Atlantic Reporter, 2d Series, be held liable for the contamination. 706). Changes are underway in common law--the 1997 Consumers do not have legal recourse if they are fully Restatement of Torts (3d) tries to eliminate the negligence aware of the products health risks but voluntarily pro- and strict liability distinction and tries to have a unified ceed to use or consume the product and are injured by approach for product liability. However, the more traditional approach from Restatement of Torts (2d) is described in this report because it is appropriate for the 1988-97 data presented in chapter 4. 4 Foresesable damages are consequences or damages that 7 The Restatement (Third) uses an example of a 1-inch an ordinary person would reasonable expect to occur, and chicken bone in a chicken enchilada, raising the question of foreseeable plaintiffs are plaintiffs than an ordinary person whether this is a manufacturing defect or an inherent aspect might expect to be potential plaintiffs. of the product (see Steenson, 1998).

36 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA foreseeable users of all foreseeable dangers, (2) the Law of Products Liability, 1987, p. 6). Sellers are not defendant failed to perform this duty, and (3) the expected to open sealed containers and inspect the defendants failure to perform this duty caused the food although some courts have stated the contrary plaintiffs injury (Harl, 1997). (American Law of Products Liability, 1987, p. 6). In more recent case law, liability through negligence Failure to warn consumers of a products hazards or of has been extended beyond manufacturers to middle- a potentially dangerous condition can result in negli- men such as distributors or warehouses, although rela- gence claims, particularly when the law requires warn- tively few cases have been brought against them ings or labels yet they were not used. In certain prod- because of the difficulty of establishing proximate ucts, manufacturers are legally obligated to provide cause (Harl, 1997). In most instances, retailers are not warning labels to alert consumers about potential dan- liable under negligence theory for latent defects unless gers. Common warnings are for foreign objects in they represent themselves as the manufacturer, assem- foods (e.g., the product may contain bones, shell, ble the package prior to its sale (Harl, 1997), or were pits, etc.). More recently, there are mandated label in a situation where they could have tested the product requirements for safe handling and cooking instruc- for safety. Food sellers are expected to use reasonable tions on raw meat as well as for refrigeration instruc- care to inspect the food that they sell and may be tion on smoked seafood packaged using modified- found negligent if there is a feasible procedure for atmosphere techniques (e.g., keep refrigerated at 38° F inspecting the food but they failed to use it (American or less). Manufacturers or product sellers may also be

Appendix table 1—Three causes of action for product liability cases

Cause of action Focus Type of law1 Description

Strict product liability Product defect Tort To prove a strict product liability action in most jurisdictions, the plaintiff must prove that the product was both defective and “unreasonably dangerous” but need not prove negligence.

Negligence Manufacturer’s Tort Either through neglect or carelessness, or seller’s conduct, the defendant failed to exercise reasonable care to prevent the product from becoming defective and harming the user.

Negligence per se Defendants violate a Federal, State, or local statute or regulation that was specifically designed to prevent the type of injuries that the plaintiff suffered.

Breach of warranty Whether product Commercial According to the Uniform Commercial Code (UCC), conforms to just by selling a product, seller incurs obligations under warranty both implied and express warranties. The plaintiff can recover damages if the product did not conform to one of these types of warranties and was not “reasonably safe.”

Breach of express warranty The representations or promises relating to the material facts (e.g., safety and wholesomeness) about the products, as described in salespersons’ statements, in pictures or writing on food containers, and in advertisements induced the consumer to buy the food. A breach occurs when these representations are not true.

Breach of implied warranty Implied warranty of merchantability “The product sold is reasonably fit for the general purpose for which it is manufactured and sold” (Harl, 1997, p. 7-18).

Implied warranty of fitness The seller knows the buyer’s intended purpose or use of the product and the buyer relies upon the seller’s judgment or skill in selecting a suitable product for the purpose.

1Torts are private or civil wrongs resulting in property damage or personal injury. Source: Adapted from Buzby and Frenzen (1999) with permission from Elsevier Press.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 37 required by law to provide health warnings for raw dards, or procedures (Rosenbaum, 1998). For exam- milk and unpasteurized juices and for restaurant and ple, in the 1993 outbreak associated with E. coli retail sale of raw shellfish. O157:H7 contaminated hamburgers, microbial standards and control procedures were established but not A related cause of action is negligence per se which effectively implemented (Rosenbaum, 1998). In par- occurs when defendants violate a statute or regulation ticular, the outbreak occurred because some of the fast- that was specifically designed to prevent the type of food restaurants failed to follow State-required cooking injury the plaintiff suffered. Negligence per se is par- procedures (Tansey, 1993). Unfortunately, compliance ticularly relevant to the food industry because there are with food safety laws and regulations does not guaran- Hazard Analysis and Critical Control Points (HACCP) tee that the food will be free of contamination. regulations and/or guidance documents that specify or suggest prevention techniques to follow. These regulations and guidance documents currently cover a broad Breach of Express or range of areas including meat and poultry, food service Implied Warranty establishments, unpasteurized fruit juices, and fish and fishery products. HACCP regulations require formal- Under the Uniform Commercial Code (UCC), sellers ized food safety and sanitation programs to be imple- incur obligations called warranties when they sell a mented with supporting documentation that they are product. Breach of warranty is a cause of action that being followed. may be claimed in a foodborne illness lawsuit and applies when the food does not conform to either an A food firm is at risk of a negligence per se claim if its express warranty or an implied warranty. In essence, food products or food service activities are covered by plaintiffs can recover compensation if the food did not a HACCP regulation but the firm has no HACCP plan conform to a warranty and that non-conforming feature in place or does not follow its own written rules, stan- of the product caused the plaintiffs injury.

Appendix figure1 Schematic of the main causes of action for product liability for food safety issues

Product liability

Breach of Strict product Negligence warranty liability

Negligence Breach of Breach of per se express warranty implied warranty

Implied warranty of Implied warranty merchantability of fitness

38 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA An express warranty is an affirmation of fact or determine what the plaintiff should have reasonably promise about the food, as made in sales respresenta- expected to be in the food. tives statements or in pictures or writing on food con- The second test is the foreign-natural test based on the tainers, menus, or advertisements, that induces the con- characteristics of the foreign material that caused the sumer to buy the food. The warranty is breached if injury. Under this test, consumers can recover from these representations prove to be false. For example, a injuries caused by an unexpected foreign substance in company that advertises its eggs as Salmonella-free the food (glass in chicken soup) but not if the same when they are not has breached an express warranty. injury was caused by a natural material in the food For food products, express warranties that are com- (bone in chicken soup). monly breached involve foreign objects in the food (bones for example). An implied warranty requires food to be both mer- Common Defenses chantable and fit for consumption. in a Product Liability Lawsuit l A product is merchantable if it meets certain pre- Issues in product liability lawsuits are usually decided scribed safety standards and is fit for the ordinary in two stages: determination of liability, and assess- purpose for which it was sold. For example, raw pork ment of damages if liability is found. The key issue in is merchantable because its ordinary purpose is human determining liability is whether or not plaintiffs can consumption after thorough cooking to kill foodborne prove causation, that is, link their foodborne illness to parasites such as Trichinae. the implicated food, to a negligent act or omission by the defendant, or to a defect or non-conforming feature l A product must also meet an implied warranty of fit- of the implicated food. ness for a particular purpose in certain representations made about it. In essence, a seller makes an implied Defendants may try to weaken the alleged causal link warranty of fitness when he knows the buyers intend- between the food and the plaintiffs illness by provided purpose or use of the product, and the buyer relies ing evidence that the illness-causing pathogen can be on the sellers judgment or skill in selecting a suitable associated with multiple foods or could have been product. For example, if a seller told a buyer that a spread via other routes, such as person-to-person con- particular type of raw fish could be used to make sushi tact. The further removed a defendant is from the and the buyer became ill from consuming the sushi plaintiff in the chain of food production, processing, because it was contaminated with Anisakid parasites, and distribution, the more difficult it may be for the this warranty would be breached. plaintiff to establish a causal link. A causal link may be easier to establish between divisions of a vertically The plaintiff is not required to show that the food sell- integrated company and harder to establish if con- er was negligent or at fault to recover damages under sumers or intervening parties such as middlemen could the implied warranty theory. Instead, the plaintiff must also have made food-handling errors. For example, in only prove that the seller sold the non-conforming very rare instances, defendants may claim that the food and that this non-conforming feature of the food foodborne illness due to home-cooked food was due to caused the plaintiffs illness. This seemingly simple the consumers faulty food-handling and preparation proof is complicated by the issues described below. practices. They may also claim that no one else Two main tests are used to determine if a food is fit for became ill while eating the same food, the food is not human consumption and whether there has been a a commonly recognized vehicle for the pathogen, or breach of implied warranty of merchantability. The that the timing of the illness is inconsistent with the most commonly adopted test is the reasonable expecta- pathogens incubation period. tion test. This is based upon what an ordinary con- Defendants may try to prove that the plaintiff was fully sumer might reasonably expect to be present in the aware of the health risks but nevertheless voluntarily food (Rubin and Lamb, 1993). For example, would proceeded to use or consume the food and then became ordinary consumers be aware of the risks from ill. Defendants may also use a plaintiffs preexisting Salmonella (Salmonella serotype Enteritidis in particu- medical condition to try to show that a plaintiff was lar) when they ate lightly poached eggs? Under this negligent or reckless. The effect of such counterclaims test, the socioeconomic, cultural, and demographic on case outcomes is unclear. characteristics of the individual plaintiff are used to

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 39 In general, if the defendant is found liable, damages Defendants found liable in foodborne illness cases may may be decreased if the defendant: (1) showed that also be assessed punitive damages if the defendant was reasonable care was taken when producing, han- found grossly negligent with food production, handling, and selling the implicated food, (2) used state- dling, processing, and sanitation practices and/or will- of-the-art technology in producing the food, and (3) fully contaminated the food. Punitive damages are followed laws and regulations designed to prevent the assessed to punish defendants for their conduct and are harm suffered by the plaintiff. A defendant that uses based upon the defendants ability to pay, not upon the its HACCP records to show that it had exercised all damages suffered by the plaintiff. However, courts reasonable preventive controls following regulatory seldom award punitive damages (Cooter, 1991). guidelines will have a better defense than a firm that Punitive damage claims do, however, have real legal cannot (Weddig, 1994). tactical use in the gathering of information during a lawsuit (discovery process) against firms and in lever- aging settlements (Clark, 2000). Also, individuals who Legal Compensation have intentionally violated a Federal or State law could for Foodborne Illness be subject to criminal prosecution which may lead to If a defendant in a foodborne illness jury trial is found additional financial and personal costs such as fines liable and compensation is awarded to the plaintiff, the and/or imprisonment (Cooter, 1991). award may be based upon components such as medical costs (past, present, and projected, after insurance), Discussion lost productivity (e.g., actual salary losses and projected lost future earnings), other dollar losses (e.g., burial Historically, many attorneys were reluctant to take expenses, travel costs to obtain medical care), and gen- foodborne illness cases because they did not know eral losses that are difficult to monetize (e.g., pain and how to proceed with them because there is so little suffering, loss of consortium (i.e., a spouses help and precedent (Rosenbaum, 1998). More recently, there affection), disability, psychological and emotional dis- has been a system-wide increase in the level of exper- tress). tise in trying these cases, and more attorneys are will- ing to accept them (Rosenbaum, 1998). In general, If there are multiple defendants, joint and several lia- attorneys with less experience with these cases are less bility might be used to allocate the burden of compen- likely to be successful (Rosenbaum, 1998). sating the plaintiff. Historically, if more than one defendant was found joint and severally liable for a Meanwhile the medical community is becoming more single indivisible injury but only one was capable of educated about foodborne disease (Rosenbaum, 1998), paying the monetary damages, that defendant might and there is increased government surveillance to iden- have to pay all of the damages. However, monetary tify foodborne illness cases, hospitalizations, and damages are commonly based upon the percent of lia- deaths. Additionally, new technologies are being bility attributed to each defendant. The legal theory of developed that can detect foodborne illnesses and contributory negligence apportions liability to the pathogens in food (e.g., Pulsed-Field Gel plaintiffs. In some jurisdictions, if the plaintiff is Electrophoresis for examining DNA fingerprints of found to be over 50 percent liable for his or her injury, specific pathogens). Greater information about food- the defendant will not have to pay any damages (Sims, borne illness and greater documentation of sporadic 1995). and outbreak cases will assist attorneys in foodborne illness lawsuits. Defendant Costs Class Action Suits In addition to any compensation paid to plaintiffs, monetary losses to a corporate defendant from an inci- A class action suit is a lawsuit brought by representa- dent resulting in a lawsuit may include any or all of the tive member(s) of a large group of persons on behalf following costs: direct and indirect costs of a product of all of the members of the group (Gifis, 1975). recall (including lost product and replacement of prod- Class action suits began to proliferate in the 1970s uct), lost profits, lost sales (i.e., affected products and (Mergenhagen, 1995). Courts may agree to accept other product in the same product line or similar prod- class action lawsuits in cases involving a large number uct category), loss of customers, reduced market share, of plaintiffs with similar damages when it is more effi- loss of business reputation, and legal expenses. cient to treat the plaintiffs as a group than as individu-

40 Product Liability and Microbial Foodborne Illness / AER-799 Economic Research Service/USDA als. However, Federal and State courts are generally A large foodborne illness outbreak due to Salmonella- unreceptive to requests to certify personal injury contaminated milk linked to Jewel Food Stores in the claims as class actions because the injuries are rarely Chicago area during the late 1980s was apparently the similar enough to justify group treatment (Clark, first class action lawsuit involving foodborne illness. 2000). Some well-known examples of consumer prod- Several other foodborne illness outbreaks since then ucts that have been involved in class action suits have also resulted in class action lawsuits. The include Dalkon shield contraceptive devices, silicone increase in class actions may reflect the widespread breast implants, and tobacco products. media coverage of the successful class action lawsuit involving the 1993 outbreak due to E. coli O157:H7- Class action lawsuits involving foodborne illness have contaminated hamburgers linked to the Jack-In-The- been rare. Severe foodborne illnesses are likely to Box restaurant chain. Class action lawsuits involving involve a variety of different symptoms and damages, foodborne illness might become even more common in particularly when chronic sequelae occur. Therefore, the future as methods for detecting outbreaks continue injuries of this kind are unlikely to meet the general to improve, and as lawyers and consumers become standard for a class action lawsuit, even in the case of more aware of the class action option for recovering large outbreaks where liability is not contested (Clark, damages due to mild illnesses associated with mass 2000). Class action lawsuits involving foodborne ill- outbreaks of foodborne illness. ness are probably most likely to be certified when a foodborne illness outbreak resulted in many mild illnesses with relatively uncomplicated claims for monetary damages.

Economic Research Service/USDA Product Liability and Microbial Foodborne Illness / AER-799 41